|
71997
|
1740
|
5
|
2026-04-22T14:39:54.980170+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868794980_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
879278402273496789
|
7411094662693935641
|
click
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP...
|
71995
|
|
71998
|
1741
|
5
|
2026-04-22T14:40:26.454680+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868826454_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-3496145305894184997
|
-1812277374160858535
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style...
|
71996
|
|
71999
|
1740
|
6
|
2026-04-22T14:40:27.424965+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868827424_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
1800618337163706643
|
-1234549984462237095
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions...
|
NULL
|
|
72000
|
1741
|
6
|
2026-04-22T14:40:58.524420+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868858524_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.
Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.
Happy to help draft a dedicated Security & Compliance section i...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI sales assistant + CRM","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strong CRM linkage","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less focus on external knowledge layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Planhat -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Related content","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ask Jiminny Anything on Call level","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Hubspot app","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Product Vision & Strategy","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Automated CRM Filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simona Georgieva (Deactivated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI agent ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steliyan Georgiev","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Comments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"James Graham","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"9 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A few areas around Security & Compliance that need more depth before this moves beyond draft:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Happy to help draft a dedicated Security & Compliance section if useful.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"15 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Expand","depth":21,"bounds":{"left":0.46991357,"top":0.19313647,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Listen","depth":21,"bounds":{"left":0.47523272,"top":0.3367917,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Show more options","depth":20,"bounds":{"left":0.47523272,"top":0.3312051,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Show thinking","depth":26,"bounds":{"left":0.3828125,"top":0.33918595,"width":0.030917553,"height":0.014764565},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXHeading","text":"Gemini said","depth":20,"bounds":{"left":0.3804854,"top":0.38068634,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gemini said","depth":21,"bounds":{"left":0.3804854,"top":0.38268158,"width":0.04105718,"height":0.01915403},"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-6004504568025778738
|
7124130923870179929
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.
Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.
Happy to help draft a dedicated Security & Compliance section i...
|
NULL
|
|
72001
|
1740
|
7
|
2026-04-22T14:40:58.939428+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868858939_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.
Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.
Happy to help draft a dedicated Security & Compliance section i...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI sales assistant + CRM","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strong CRM linkage","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less focus on external knowledge layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Planhat -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Related content","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ask Jiminny Anything on Call level","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Hubspot app","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Product Vision & Strategy","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Automated CRM Filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simona Georgieva (Deactivated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI agent ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steliyan Georgiev","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Comments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"James Graham","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"9 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A few areas around Security & Compliance that need more depth before this moves beyond draft:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Happy to help draft a dedicated Security & Compliance section if useful.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"15 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Expand","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Listen","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Show more options","depth":20,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Show thinking","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXHeading","text":"Gemini said","depth":20,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gemini said","depth":21,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Thank you for sharing the draft PRD for the","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":". It looks like a highly strategic initiative to transition Jiminny from a standalone platform into a foundational data layer for AI-driven revenue workflows.","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Here is a quick breakdown of the document as it currently stands:","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXHeading","text":"Executive Summary","depth":23,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Executive Summary","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Objective:","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an MCP-based AI Knowledge Connector to expose Jiminny data (transcripts, summaries, deal context) to external AI tools like Claude, OpenAI, and Gemini.","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Target Value:","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increases product stickiness, reduces manual data entry/copy-pasting, and achieves competitive parity with platforms like Gong and Avoma.","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Capabilities:","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"* Call and deal-level data retrieval.","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Semantic and keyword search.","depth":28,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built prompts, workflows, and skills (e.g., deal inspection, objection analysis).","depth":28,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Current Bottlenecks & Discussions:","depth":26,"bounds":{"left":0.23159721,"top":0.0,"width":0.19097222,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"James Graham has flagged several critical security and compliance considerations (auth modeling, data minimization, key lifecycle, rate limiting, and GDPR downstream processing). Galya has integrated some of these into the technical requirements but pushed back on the GDPR point, noting that customers already export this data manually or via standard APIs.","depth":26,"bounds":{"left":0.23159721,"top":0.0,"width":0.21875,"height":0.31166667},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Since you shared the document without a specific prompt, how would you like to proceed—would you like help drafting a response to Galya's GDPR question, refining the technical requirements based on James's feedback, or summarizing this into a pitch for stakeholders?","depth":24,"bounds":{"left":0.20729166,"top":0.25111112,"width":0.2375,"height":0.16722222},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXTextArea","text":"Enter a prompt for Gemini\nencrypted","depth":20,"value":"Enter a prompt for Gemini\nencrypted","help_text":"","role_description":"text entry area","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Enter a prompt for Gemini","depth":21,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"encrypted","depth":21,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Open upload file menu","depth":20,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Tools","depth":18,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Open mode picker","depth":20,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false}]...
|
-6004504568025778738
|
7124130923870179929
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.
Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.
Happy to help draft a dedicated Security & Compliance section i...
|
71999
|
|
72002
|
1741
|
7
|
2026-04-22T14:41:29.623928+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868889623_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-5478279924297120505
|
-2099241112917355943
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring...
|
72000
|
|
72003
|
1740
|
8
|
2026-04-22T14:41:30.331396+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868890331_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-5142066327100965622
|
-2099241112917484967
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences...
|
NULL
|
|
72004
|
1741
|
8
|
2026-04-22T14:42:00.767444+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868920767_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.
Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.
Happy to help draft a dedicated Security & Compliance section i...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI sales assistant + CRM","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strong CRM linkage","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less focus on external knowledge layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Planhat -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Related content","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ask Jiminny Anything on Call level","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Hubspot app","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Product Vision & Strategy","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Automated CRM Filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simona Georgieva (Deactivated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI agent ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steliyan Georgiev","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Comments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"James Graham","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"9 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A few areas around Security & Compliance that need more depth before this moves beyond draft:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Happy to help draft a dedicated Security & Compliance section if useful.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-6004504568025778738
|
7124130923870179929
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.
Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.
Happy to help draft a dedicated Security & Compliance section i...
|
NULL
|
|
72005
|
1740
|
9
|
2026-04-22T14:42:01.606287+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868921606_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
8435392526505937761
|
-2099241112917484967
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility...
|
72003
|
|
72006
|
1741
|
9
|
2026-04-22T14:42:31.999168+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868951999_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-468960449895495518
|
7412220562600647257
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow...
|
72004
|
|
72007
|
1740
|
10
|
2026-04-22T14:42:32.679367+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776868952679_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
8709169268692632247
|
-2099241112917484967
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails...
|
NULL
|
|
72032
|
1743
|
9
|
2026-04-22T14:46:24.678247+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869184678_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-6119790967225190804
|
7421227770445452825
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude...
|
72023
|
|
72089
|
1745
|
17
|
2026-04-22T14:51:51.949449+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869511949_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-4610773300625247343
|
-1812136636739594727
|
app_switch
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH...
|
NULL
|
|
72090
|
1744
|
17
|
2026-04-22T14:51:52.441451+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869512441_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-7954132901822255868
|
7412220571190712857
|
app_switch
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly...
|
NULL
|
|
72091
|
1744
|
18
|
2026-04-22T14:52:02.885614+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869522885_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
168412454662634921
|
-1812206455996411367
|
click
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools...
|
72090
|
|
72092
|
1745
|
18
|
2026-04-22T14:52:03.284454+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869523284_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-3226341313328606922
|
7412220562600762905
|
click
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH...
|
72089
|
|
72093
|
1744
|
19
|
2026-04-22T14:52:35.402854+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869555402_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
6879718307218968081
|
-2099241112917355943
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain...
|
NULL
|
|
72094
|
1745
|
19
|
2026-04-22T14:52:36.130317+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869556130_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.
Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.
Happy to help draft a dedicated Security & Compliance section i...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI sales assistant + CRM","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strong CRM linkage","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less focus on external knowledge layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Planhat -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Related content","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ask Jiminny Anything on Call level","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Hubspot app","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Product Vision & Strategy","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Automated CRM Filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simona Georgieva (Deactivated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI agent ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steliyan Georgiev","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Comments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"James Graham","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"9 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A few areas around Security & Compliance that need more depth before this moves beyond draft:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Happy to help draft a dedicated Security & Compliance section if useful.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"15 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Expand","depth":21,"bounds":{"left":0.46991357,"top":0.19313647,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Listen","depth":22,"bounds":{"left":0.47523272,"top":0.3367917,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Show more options","depth":20,"bounds":{"left":0.47523272,"top":0.3312051,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Show thinking","depth":26,"bounds":{"left":0.3828125,"top":0.33918595,"width":0.030917553,"height":0.014764565},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXHeading","text":"Gemini said","depth":20,"bounds":{"left":0.3804854,"top":0.38068634,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gemini said","depth":21,"bounds":{"left":0.3804854,"top":0.38268158,"width":0.04105718,"height":0.01915403},"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-6004504568025778738
|
7124130923870179929
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.
Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.
Happy to help draft a dedicated Security & Compliance section i...
|
NULL
|
|
72095
|
NULL
|
0
|
2026-04-22T14:53:06.509562+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869586509_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
6041603874425459045
|
-2099241112917372327
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context...
|
72093
|
|
72096
|
NULL
|
0
|
2026-04-22T14:53:07.265864+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869587265_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-270178233906525078
|
7412361299820565081
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name...
|
72094
|
|
72097
|
1746
|
0
|
2026-04-22T14:53:37.585887+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869617585_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-6545360246787818106
|
-1811010736765660583
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period...
|
NULL
|
|
72098
|
1747
|
0
|
2026-04-22T14:53:38.417861+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776869618417_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.
Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.
Happy to help draft a dedicated Security & Compliance section i...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.48922586,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI sales assistant + CRM","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strong CRM linkage","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less focus on external knowledge layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Planhat -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Related content","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ask Jiminny Anything on Call level","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Hubspot app","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Product Vision & Strategy","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Automated CRM Filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simona Georgieva (Deactivated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI agent ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steliyan Georgiev","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Comments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"James Graham","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"9 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A few areas around Security & Compliance that need more depth before this moves beyond draft:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Happy to help draft a dedicated Security & Compliance section if useful.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"15 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Expand","depth":21,"bounds":{"left":0.46991357,"top":0.19313647,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Listen","depth":22,"bounds":{"left":0.47523272,"top":0.3367917,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Show more options","depth":20,"bounds":{"left":0.47523272,"top":0.3312051,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Show thinking","depth":26,"bounds":{"left":0.3828125,"top":0.33918595,"width":0.030917553,"height":0.014764565},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXHeading","text":"Gemini said","depth":20,"bounds":{"left":0.3804854,"top":0.38068634,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gemini said","depth":21,"bounds":{"left":0.3804854,"top":0.38268158,"width":0.04105718,"height":0.01915403},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Thank you for sharing the draft PRD for the","depth":24,"bounds":{"left":0.3695146,"top":0.39026338,"width":0.10521942,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":24,"bounds":{"left":0.3695146,"top":0.41101357,"width":0.0625,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":". It looks like a highly strategic initiative to transition Jiminny from a standalone platform into a foundational data layer for AI-driven revenue workflows.","depth":24,"bounds":{"left":0.3695146,"top":0.41101357,"width":0.11319814,"height":0.07861133},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Here is a quick breakdown of the document as it currently stands:","depth":24,"bounds":{"left":0.3695146,"top":0.5067837,"width":0.11635638,"height":0.037110932},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXHeading","text":"Executive Summary","depth":23,"bounds":{"left":0.3695146,"top":0.56823623,"width":0.11635638,"height":0.01915403},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Executive Summary","depth":24,"bounds":{"left":0.3695146,"top":0.5698324,"width":0.051363032,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Objective:","depth":26,"bounds":{"left":0.38115028,"top":0.5961692,"width":0.040392287,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an MCP-based AI Knowledge Connector to expose Jiminny data (transcripts, summaries, deal context) to external AI tools like Claude, OpenAI, and Gemini.","depth":26,"bounds":{"left":0.38115028,"top":0.5961692,"width":0.10472074,"height":0.09936153},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Target Value:","depth":26,"bounds":{"left":0.38115028,"top":0.70630485,"width":0.03374335,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increases product stickiness, reduces manual data entry/copy-pasting, and achieves competitive parity with platforms like Gong and Avoma.","depth":26,"bounds":{"left":0.38115028,"top":0.70630485,"width":0.10305851,"height":0.07861133},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Capabilities:","depth":26,"bounds":{"left":0.38115028,"top":0.79569036,"width":0.043550532,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"* Call and deal-level data retrieval.","depth":26,"bounds":{"left":0.38115028,"top":0.79569036,"width":0.09458112,"height":0.037110932},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Semantic and keyword search.","depth":28,"bounds":{"left":0.39145613,"top":0.8435754,"width":0.073803194,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built prompts, workflows, and skills (e.g., deal inspection, objection analysis).","depth":28,"bounds":{"left":0.39145613,"top":0.8707103,"width":0.08976064,"height":0.057861134},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Current Bottlenecks & Discussions:","depth":26,"bounds":{"left":0.38115028,"top":0.9393456,"width":0.09142287,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"James Graham has flagged several critical security and compliance considerations (auth modeling, data minimization, key lifecycle, rate limiting, and GDPR downstream processing). Galya has integrated some of these into the technical requirements but pushed back on the GDPR point, noting that customers already export this data manually or via standard APIs.","depth":26,"bounds":{"left":0.38115028,"top":0.9393456,"width":0.10472074,"height":0.0606544},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Since you shared the document without a specific prompt, how would you like to proceed—would you like help drafting a response to Galya's GDPR question, refining the technical requirements based on James's feedback, or summarizing this into a pitch for stakeholders?","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXTextArea","text":"Enter a prompt for Gemini\nencrypted","depth":20,"bounds":{"left":0.37416887,"top":0.8104549,"width":0.10638298,"height":0.01915403},"value":"Enter a prompt for Gemini\nencrypted","help_text":"","role_description":"text entry area","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Enter a prompt for Gemini","depth":21,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"encrypted","depth":21,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Open upload file menu","depth":20,"bounds":{"left":0.37017953,"top":0.8447725,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Tools","depth":18,"bounds":{"left":0.38613698,"top":0.8447725,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Open mode picker","depth":20,"bounds":{"left":0.4431516,"top":0.8439745,"width":0.026097074,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pro","depth":23,"bounds":{"left":0.44847074,"top":0.8527534,"width":0.007480053,"height":0.014764565},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXCheckBox","text":"Microphone","depth":19,"bounds":{"left":0.47124335,"top":0.8439745,"width":0.013297873,"height":0.031923383},"role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false}]...
|
-6004504568025778738
|
7124130923870179929
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.
Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.
Happy to help draft a dedicated Security & Compliance section i...
|
NULL
|
|
72276
|
1756
|
11
|
2026-04-22T15:22:09.775597+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871329775_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-5541539164979447414
|
-1811010737034096039
|
visual_change
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot...
|
72274
|
|
72277
|
1757
|
19
|
2026-04-22T15:22:13.265290+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871333265_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-8388431772346876836
|
-1811010187345374631
|
app_switch
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context...
|
NULL
|
|
72278
|
1757
|
20
|
2026-04-22T15:22:19.241558+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871339241_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-2837684395533528161
|
-2099241113185791399
|
visual_change
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question...
|
72277
|
|
72279
|
1756
|
12
|
2026-04-22T15:22:44.924869+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871364924_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's acti...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI sales assistant + CRM","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strong CRM linkage","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less focus on external knowledge layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Planhat -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Related content","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ask Jiminny Anything on Call level","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Hubspot app","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Product Vision & Strategy","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Automated CRM Filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simona Georgieva (Deactivated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI agent ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steliyan Georgiev","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Comments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"James Graham","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"9 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A few areas around Security & Compliance that need more depth before this moves beyond draft:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Happy to help draft a dedicated Security & Compliance section if useful.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"15 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Expand","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Listen","depth":22,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Show more options","depth":20,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Show thinking","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXHeading","text":"Gemini said","depth":20,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gemini said","depth":21,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Thank you for sharing the draft PRD for the","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":". It looks like a highly strategic initiative to transition Jiminny from a standalone platform into a foundational data layer for AI-driven revenue workflows.","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Here is a quick breakdown of the document as it currently stands:","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXHeading","text":"Executive Summary","depth":23,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Executive Summary","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Objective:","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an MCP-based AI Knowledge Connector to expose Jiminny data (transcripts, summaries, deal context) to external AI tools like Claude, OpenAI, and Gemini.","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Target Value:","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increases product stickiness, reduces manual data entry/copy-pasting, and achieves competitive parity with platforms like Gong and Avoma.","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
1182909288348349150
|
-2099240563497086375
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's acti...
|
NULL
|
|
72280
|
1757
|
21
|
2026-04-22T15:22:50.743936+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871370743_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples....
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
5637940384698117010
|
-4116853746247773607
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples....
|
NULL
|
|
72281
|
1756
|
13
|
2026-04-22T15:23:16.003200+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871396003_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-292897919630028625
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server...
|
72279
|
|
72282
|
1757
|
22
|
2026-04-22T15:23:21.919562+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871401919_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
9021723979828606460
|
7412220562332340761
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low...
|
72280
|
|
72283
|
NULL
|
0
|
2026-04-22T15:23:47.164348+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871427164_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's acti...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI sales assistant + CRM","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strong CRM linkage","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less focus on external knowledge layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Planhat -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Related content","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ask Jiminny Anything on Call level","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Hubspot app","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Product Vision & Strategy","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Automated CRM Filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simona Georgieva (Deactivated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI agent ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steliyan Georgiev","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Comments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"James Graham","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"9 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A few areas around Security & Compliance that need more depth before this moves beyond draft:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Happy to help draft a dedicated Security & Compliance section if useful.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"15 Apr","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Expand","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Listen","depth":22,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Show more options","depth":20,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Show thinking","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXHeading","text":"Gemini said","depth":20,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gemini said","depth":21,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Thank you for sharing the draft PRD for the","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":". It looks like a highly strategic initiative to transition Jiminny from a standalone platform into a foundational data layer for AI-driven revenue workflows.","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Here is a quick breakdown of the document as it currently stands:","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXHeading","text":"Executive Summary","depth":23,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Executive Summary","depth":24,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Objective:","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an MCP-based AI Knowledge Connector to expose Jiminny data (transcripts, summaries, deal context) to external AI tools like Claude, OpenAI, and Gemini.","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Target Value:","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increases product stickiness, reduces manual data entry/copy-pasting, and achieves competitive parity with platforms like Gong and Avoma.","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Capabilities:","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"* Call and deal-level data retrieval.","depth":26,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Semantic and keyword search.","depth":28,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built prompts, workflows, and skills (e.g., deal inspection, objection analysis).","depth":28,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Current Bottlenecks & Discussions:","depth":26,"bounds":{"left":0.23159721,"top":0.0,"width":0.19097222,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"James Graham has flagged several critical security and compliance considerations (auth modeling, data minimization, key lifecycle, rate limiting, and GDPR downstream processing). Galya has integrated some of these into the technical requirements but pushed back on the GDPR point, noting that customers already export this data manually or via standard APIs.","depth":26,"bounds":{"left":0.23159721,"top":0.0,"width":0.21875,"height":0.31166667},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Since you shared the document without a specific prompt, how would you like to proceed—would you like help drafting a response to Galya's GDPR question, refining the technical requirements based on James's feedback, or summarizing this into a pitch for stakeholders?","depth":24,"bounds":{"left":0.20729166,"top":0.25111112,"width":0.2375,"height":0.16722222},"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
1182909288348349150
|
-2099240563497086375
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)
Competitive analysis
Galya Dimitrova
AI agent ideas
Steliyan Georgiev
Comments
James Graham
9 Apr
A few areas around Security & Compliance that need more depth before this moves beyond draft:
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Team visibility enforcement should be its own HIGH requirement. The current note under access control ("ensure MCP only exposes data user has access to") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped.
No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this.
GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's acti...
|
NULL
|
|
72284
|
NULL
|
0
|
2026-04-22T15:23:53.732163+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871433732_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-7182536218736414846
|
-4116853746247773607
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments...
|
NULL
|
|
72285
|
1758
|
0
|
2026-04-22T15:24:18.248337+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871458248_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
5558808362347335703
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access...
|
72283
|
|
72286
|
1759
|
0
|
2026-04-22T15:24:25.111547+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871465111_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve....
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
3815961420847836129
|
-4116853746247773607
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve....
|
72284
|
|
72287
|
1758
|
1
|
2026-04-22T15:24:49.357353+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871489357_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
1312122821506438570
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items...
|
NULL
|
|
72288
|
1759
|
1
|
2026-04-22T15:24:52.157454+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871492157_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
5050895807565877683
|
-2099241113185791399
|
visual_change
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights...
|
NULL
|
|
72289
|
1759
|
2
|
2026-04-22T15:25:01.487132+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871501487_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-7182536218736414846
|
-4116853746247773607
|
visual_change
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments...
|
72288
|
|
72290
|
1758
|
2
|
2026-04-22T15:25:20.442539+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871520442_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
2161715775331057805
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript...
|
72287
|
|
72291
|
1759
|
3
|
2026-04-22T15:25:33.266227+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871533266_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
4325390886770462967
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention...
|
NULL
|
|
72292
|
1758
|
3
|
2026-04-22T15:25:52.169042+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871552169_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
4584936107819514130
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics...
|
NULL
|
|
72293
|
1759
|
4
|
2026-04-22T15:26:04.418109+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871564418_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-1723504415601968864
|
-4116853746247773607
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope...
|
72291
|
|
72294
|
1758
|
4
|
2026-04-22T15:26:23.254665+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871583254_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-1684593866535502482
|
-4116853746247773671
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request...
|
72292
|
|
72295
|
1759
|
5
|
2026-04-22T15:26:35.740767+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871595740_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
3300266133540403640
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows...
|
NULL
|
|
72296
|
1758
|
5
|
2026-04-22T15:26:56.087313+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871616087_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
3300266133540403640
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows...
|
NULL
|
|
72297
|
1759
|
6
|
2026-04-22T15:27:07.110656+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871627110_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
41659848958013753
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis...
|
72295
|
|
72298
|
1758
|
6
|
2026-04-22T15:27:27.292813+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871647292_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
7917070406292614048
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness...
|
72296
|
|
72299
|
1759
|
7
|
2026-04-22T15:27:38.245841+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871658245_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI sales assistant + CRM","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strong CRM linkage","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less focus on external knowledge layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Planhat -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Related content","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ask Jiminny Anything on Call level","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Hubspot app","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Product Vision & Strategy","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Automated CRM Filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simona Georgieva (Deactivated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
6829082270594834736
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova
Automated CRM Filling
Simona Georgieva (Deactivated)...
|
NULL
|
|
72300
|
1758
|
7
|
2026-04-22T15:27:58.355505+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871678355_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI sales assistant + CRM","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strong CRM linkage","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less focus on external knowledge layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Planhat -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Related content","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ask Jiminny Anything on Call level","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Hubspot app","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
8417999621461651077
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app...
|
NULL
|
|
72301
|
1759
|
8
|
2026-04-22T15:28:09.343484+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871689343_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
-4485512217746663569
|
-1811010737034096039
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case...
|
72299
|
|
72302
|
1758
|
8
|
2026-04-22T15:28:29.420947+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871709420_m1.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_1
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.028819444,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.051736113,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.075,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.09826389,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.121527776,"top":0.0,"width":0.022222223,"height":0.035555556},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.30868056,"top":0.125,"width":0.077430554,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"bounds":{"left":0.30868056,"top":0.18777777,"width":0.074652776,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"bounds":{"left":0.30868056,"top":0.28111112,"width":0.10555556,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"bounds":{"left":0.30868056,"top":0.37444445,"width":0.108333334,"height":0.085},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"bounds":{"left":0.30868056,"top":0.4677778,"width":0.1,"height":0.24055555},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"bounds":{"left":0.30868056,"top":0.71666664,"width":0.0875,"height":0.116111115},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"bounds":{"left":0.30868056,"top":0.9027778,"width":0.07083333,"height":0.054444443},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"bounds":{"left":0.30868056,"top":0.96555555,"width":0.072222225,"height":0.022777777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"bounds":{"left":0.30868056,"top":0.99666667,"width":0.099305555,"height":0.0033333302},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"bounds":{"left":0.30868056,"top":1.0,"width":0.08715278,"height":-0.09000003},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
2997014496364196567
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations...
|
72300
|
|
72303
|
NULL
|
0
|
2026-04-22T15:28:41.373686+00:00
|
/Users/lukas/.screenpipe/data/data/2026-04-22/1776 /Users/lukas/.screenpipe/data/data/2026-04-22/1776871721373_m2.jpg...
|
Firefox
|
Jiminny MCP Connector - Product - Confluence — Wor Jiminny MCP Connector - Product - Confluence — Work...
|
True
|
jiminny.atlassian.net/wiki/spaces/PROD/pages/37282 jiminny.atlassian.net/wiki/spaces/PROD/pages/3728244737/Jiminny+MCP+Connector...
|
monitor_2
|
NULL
|
NULL
|
NULL
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova...
|
[{"role":"AXRadioButton","text [{"role":"AXRadioButton","text":"Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira","depth":4,"bounds":{"left":0.28307846,"top":0.0518755,"width":0.07596409,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":4,"bounds":{"left":0.28125,"top":0.09497207,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"SevenShores\\Hubspot\\Exceptions\\BadRequest: Client error: `POST https://api.hubapi.com/crm/v3/objects/contact/search` resulted in a `429 Too Many Requests` response: {\"status\":\"error\",\"message\":\"You have reached your secondly limit.\",\"errorType\":\"RATE_LIMIT","depth":5,"bounds":{"left":0.2945479,"top":0.10614525,"width":0.4644282,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":4,"bounds":{"left":0.28125,"top":0.12769353,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20372] AI Reports > Empty page design and promotion - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.13886672,"width":0.11319814,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.16041501,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.17158818,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.19313647,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.20430966,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Workers | Datadog","depth":4,"bounds":{"left":0.28125,"top":0.22585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Workers | Datadog","depth":5,"bounds":{"left":0.2945479,"top":0.23703113,"width":0.032081116,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Pull requests · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.2585794,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Pull requests · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.2697526,"width":0.04537899,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":4,"bounds":{"left":0.28125,"top":0.29130086,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"[JY-20728] [HubSpot] Find the root cause of 429 hit and tweak API client rate limiter - Jira","depth":5,"bounds":{"left":0.2945479,"top":0.30247405,"width":0.15791224,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.32402235,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.33519554,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny","depth":4,"bounds":{"left":0.28125,"top":0.3567438,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Jiminny","depth":5,"bounds":{"left":0.2945479,"top":0.367917,"width":0.013131649,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"CloudWatch | us-east-2","depth":4,"bounds":{"left":0.28125,"top":0.38946527,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"CloudWatch | us-east-2","depth":5,"bounds":{"left":0.2945479,"top":0.40063846,"width":0.041223403,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":4,"bounds":{"left":0.28125,"top":0.42218676,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Your 'Not enpough activities' report wasn't generated - lukas.kovalik@jiminny.com - Jiminny Mail","depth":5,"bounds":{"left":0.2945479,"top":0.43335995,"width":0.16821809,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Jiminny MCP Connector - Product - Confluence","depth":4,"bounds":{"left":0.28125,"top":0.45490822,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":true},{"role":"AXStaticText","text":"Jiminny MCP Connector - Product - Confluence","depth":5,"bounds":{"left":0.2945479,"top":0.4660814,"width":0.08294548,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Close tab","depth":5,"bounds":{"left":0.34857047,"top":0.46209097,"width":0.007978723,"height":0.01915403},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXRadioButton","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.48762968,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | change nudges schema by nikolaybiaivanov · Pull Request #11983 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.49880287,"width":0.15475398,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":4,"bounds":{"left":0.28125,"top":0.5203512,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"JY-9712 | Nuges to expire after one year by nikolaybiaivanov · Pull Request #11981 · jiminny/app","depth":5,"bounds":{"left":0.2945479,"top":0.53152436,"width":0.16555852,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Architecture overview - Model Context Protocol","depth":4,"bounds":{"left":0.28125,"top":0.55307263,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Architecture overview - Model Context Protocol","depth":5,"bounds":{"left":0.2945479,"top":0.5642458,"width":0.08277926,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.5857941,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP - AI tooling for Laravel, by the Laravel team | Laravel - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.5969673,"width":0.18583776,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXRadioButton","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":4,"bounds":{"left":0.28125,"top":0.61851555,"width":0.07962101,"height":0.032721467},"help_text":"","role_description":"tab","subrole":"AXTabButton","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXStaticText","text":"Laravel MCP | Laravel 13.x - The clean stack for Artisans and agents","depth":5,"bounds":{"left":0.2945479,"top":0.62968874,"width":0.1178524,"height":0.010774142},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"New Tab","depth":4,"bounds":{"left":0.2840758,"top":0.6528332,"width":0.07413564,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Customize sidebar","depth":6,"bounds":{"left":0.2840758,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Close Google Gemini (⌃X)","depth":6,"bounds":{"left":0.29504654,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Tabs from other devices","depth":6,"bounds":{"left":0.30618352,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open history (⇧⌘H)","depth":6,"bounds":{"left":0.31732047,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXCheckBox","text":"Open bookmarks (⌘B)","depth":6,"bounds":{"left":0.32845744,"top":0.97007185,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"toggle button","subrole":"AXToggle","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"AI Chat settings","depth":7,"bounds":{"left":0.4659242,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Close","depth":7,"bounds":{"left":0.47789228,"top":0.055067837,"width":0.010638298,"height":0.025538707},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"WORK, Google Account: lukas.kovalik@jiminny.com","depth":12,"bounds":{"left":0.47523272,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXButton","text":"Main menu","depth":12,"bounds":{"left":0.3648604,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"New Chat","depth":12,"bounds":{"left":0.44730717,"top":0.103751,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Open menu for conversation actions.","depth":12,"bounds":{"left":0.46060506,"top":0.103751,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false,"is_expanded":false},{"role":"AXHeading","text":"Conversation with Gemini","depth":15,"bounds":{"left":0.36053857,"top":0.14764565,"width":0.0003324468,"height":0.0007980846},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Conversation with Gemini","depth":16,"bounds":{"left":0.36053857,"top":0.15003991,"width":0.1200133,"height":0.025538707},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXButton","text":"Copy prompt","depth":21,"bounds":{"left":0.3834774,"top":0.18355946,"width":0.013297873,"height":0.031923383},"help_text":"","role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXButton","text":"Edit","depth":21,"bounds":{"left":0.39810506,"top":0.18355946,"width":0.013297873,"height":0.031923383},"role_description":"button","subrole":"AXUnknown","is_enabled":true,"is_focused":false,"is_selected":false},{"role":"AXHeading","text":"You said Jiminny MCP Connector Epic Link to Epik in Jira Document status DRAFT Objective Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows. Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product. 👤 Target user Revenue teams using AI tools (Sales, CS, RevOps) Companies already experimenting with Claude / OpenAI Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc) 🤕 Pain point or problem Jiminny data is locked in the platform AI tools lack access to high-value conversation context Customers must manually export/copy transcripts No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people Peer pressure as almost all competitors have this 💥 Impact and benefits For customers Better AI outputs using real customer conversations Ability to build automated workflows across tools For Jiminny Increased product stickiness (embedded in workflows) Competitive parity with Gong / Avoma etc. 💡 Solution ideas Build an AI Knowledge Connector (MCP-based) that: Exposes Jiminny data (calls, transcripts, metadata) to AI tools Provides structured insights (summaries, action items, scoring etc.) Supports deal-level context Allows integration into customer-owned AI stacks (Claude + Zapier, etc.) Success metrics List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets). Goal Metric Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows % of customers connecting external AI tools Number of MCP/API calls per customer - and which ones are being used the most Decrease the manual copying of transcripts decrease by X % User interaction and design Simple setup flow: Generate API key / connect MCP Clear instructions for Claude / Zapier setup Provide: Example prompts for Claude Pre-built workflow templates for Zapier Clear documentation - with use cases 👉 Focus: usable by non-technical users Detailed Requirements List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification. Requirement Importance Notes Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI) HIGH Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack Allow customers to access call data through AI tools HIGH Expose Jiminny data: list calls (filter by date, user, account, deal) get call get transcript (speaker-separated) get call metadata search calls (keyword + semantic) Support filtering by: account / contact/ lead / deal jiminny user date range Allow customers to search and query calls using AI tools HIGH Support keyword and semantic search across conversations Provide structured insights so customers can generate better outputs HIGH For each call, expose: Summary Action items Key points Activity type Notes Call Score CRM information Allow customers to understand deal status across multiple calls/emails HIGH For each deal expose: calls/meetings/emails Identified risks contacts related to the deal with their job title linked account details all CRM field information that we have in DI + fields used for CRM filling example - get_deal_summary request - { \"tool\": \"get_deal_summary\", \"arguments\": { \"deal_name\": \"ACME\" } } responce: { \"deal_name\": \"ACME\", \"stage\": \"Negotiation\", \"recent_activity\": [ \"Demo call on 2026-04-01\", \"Pricing discussion on 2026-04-03\" ], \"risks\": [ \"Pricing concerns\", \"Security review pending\" ], \"next_steps\": [ \"Send follow-up to VP Sales\", \"Share security documentation\" ] } Ensure the AI can easily understand the schema of the data coming from Jiminny HIGH Models perform much better when tools have: clear names rich descriptions predictable schemas Provide clear, actionable error states for customers when MCP connections fail HIGH Error messages should guide agents toward solutions with specific suggestions and next steps. Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.” Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.” Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.” Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.” Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.” Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.” No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.” Allow customers to securely control who and what data is accessible externally HIGH Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well. Private meetings shouldn’t be exposed via MCP HIGH When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP. Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows HIGH Ensure data is structured to be easily combined with: CRM (accounts, deals) external docs Use consistent identifiers: account IDs deal IDs MCP is available only for Scale tier and above HIGH Customers who are on Capture tier should be able to access any of their data through MCP. This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working. Provide visibility into how external AI tools access Jiminny data HIGH Audit logs: what data was accessed by which external system when identity - who made the request specific queries/filters used failed access attempts (especially cross-visibility-boundary attempts) Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it. example from Atlassian - Provide customers with example prompts to get started quickly HIGH MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments. See examples in table below. Expose Jiminny as a Connector in Claude medium Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny Provide customers with example workflows to get started quickly medium A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome. See examples in table below. Allow customers to use predefined Jiminny skills for common workflows medium Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job. Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com See examples in table below. Jiminny UI widget for OpenAI medium The reason to have this is to make certain Jiminny use cases much easier than pure chat. Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it. For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface. Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section. Allow customers to run long-running analyses asynchronously and retrieve progress/results low Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to: start the job let it run in the background return a status like “in progress” let the client come back for the result later For example - User asks in Claude: “Analyse all calls from this quarter and tell me the top 5 objections by segment.” That might require: searching lots of calls pulling many summaries/transcripts aggregating results Instead of making Claude wait 45+ seconds, Jiminny could: create an analysis job return: job_id = 123, status = running We need this in order to support: large-scale analysis multi-call aggregation across lots of data playlist generation across many calls heavy recap/report generation Creating a snippet through the MCP low Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team. example request: { \"tool\": \"create_snippet\", \"arguments\": { \"call_id\": \"call_123\", \"start_time\": \"00:18:24\", \"end_time\": \"00:19:10\", \"title\": \"Pricing objection\" } } example response: { \"snippet_id\": \"snip_987\", \"title\": \"Pricing objection\", \"share_url\": \"https://app.jiminny.com/snippets/snip_987\" } Ability to add recordings or snippets to a playlist low Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes Allow AI tools to read reusable Jiminny resources low fathom Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be: current deal snapshot account summary weekly rep recap approved talk tracks / objection handling guidance curated playlists or snippet libraries This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls. Ability to write coaching notes in Jiminny through MCP low What we want to do for prompts/worflows and skills Capability Type User value / job to be done Example Jiminny use case Inputs Steps performed Output Draft follow-up email Prompt Help reps quickly create tailored follow-ups after calls Use the latest customer call to draft an email with recap and next steps Call ID optional tone optional recipient type Retrieve call summary and action items from Jiminny. Pull key points, commitments, and next steps. Apply the chosen tone and structure. Draft the email in a consistent format. Customer-ready follow-up email draft Summarise call Prompt Help users quickly extract the most important points from a conversation Generate a concise summary of one call for internal or external use Call ID optional summary style optional audience Retrieve transcript, metadata, and structured insights. Identify key topics, decisions, risks, and next steps. Format output for the chosen audience. Short or structured call summary with actions and risks Generate deal recap Prompt Help reps and managers get a fast overview of a deal Summarise the current state of a deal based on recent calls and linked CRM context Deal ID or account name optional date range Retrieve calls and meetings linked to the deal. Pull deal metadata and CRM-linked context. Identify recent changes, blockers, risks, and next steps. Summarise into a concise recap. Deal recap with status, blockers, risks, and recommended next steps Prepare weekly recap Prompt Help managers or reps review what changed over a time period Weekly recap of calls, themes, and actions for one rep, team, or account Rep ID team ID or account ID time period Retrieve relevant calls and summaries for the selected time period. Aggregate themes, action items, and changes. Highlight notable risks, opportunities, and follow-ups. Weekly recap with top themes, risks, opportunities, and actions Prepare deal review Workflow Help reps and managers review deal status using multiple sources of context Review a deal before pipeline review or forecast meeting Deal ID optional date range Retrieve all recent calls, emails, and meetings tied to the deal. Pull structured call summaries, score, action items, and key points. Pull linked CRM context such as stage, contacts, and account details. Aggregate open risks, unresolved objections, stakeholder gaps, and next steps. Format into a review-ready structure. Deal review document with current status, risks, stakeholders, blockers, and next steps Weekly team recap Workflow Help managers monitor activity and risk across a group of reps or accounts End-of-week team summary for sales leadership Team ID date range Retrieve calls and structured insights across the team. Group by rep, account, or deal. Identify recurring themes, major risks, strong calls, and missed follow-ups. Highlight calls or deals that need attention. Team-level recap with patterns, key risks, notable wins, and action areas Objection analysis across calls Workflow Help teams identify patterns across multiple conversations Analyse pricing or competitor objections across recent calls Account ID deal ID rep ID keyword/topic date range Search calls matching the topic or keyword. Retrieve summaries, transcripts, and snippets where relevant. Identify recurring objections and how they were handled. Assess whether objections were resolved or remain open. Summarise patterns and examples. Objection analysis with themes, frequency, examples, and suggested follow-up actions Build onboarding playlist Workflow Help enablement teams curate useful learning content faster Build a playlist of strong discovery, demo, or pricing calls for new hires Topic call type rep/team filter optional time period Search for calls matching the selected criteria. Rank or filter calls based on score, topics, or key moments. Identify useful snippets or full recordings. Add selected calls/snippets into a playlist structure. Generate short descriptions for each item. Curated onboarding playlist with selected calls/snippets and descriptions Deal Inspection Skill Skill Give users a consistent, opinionated way to assess a deal Claude reviews a deal using a repeatable Jiminny methodology Deal ID or account name optional review scope Determine whether the request is suitable for a deal inspection. Retrieve relevant calls, summaries, CRM context, stakeholders, and risks. Apply a standard Jiminny review framework to assess deal health. Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs. Present findings in a consistent structure with recommended actions. Structured deal inspection with health assessment, red flags, and recommendations Manager Coaching Skill Skill Help managers review calls consistently and identify coaching moments Analyse a rep’s call and surface strengths, gaps, and examples for coaching Call ID optional rep ID optional coaching focus Retrieve transcript, score, summary, and key moments from the call. Evaluate the conversation against a coaching framework. Identify what went well and where the rep could improve. Suggest moments to clip as snippets. Recommend coaching actions or playlist additions. Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions CSM / QBR Prep Skill Skill Help CS or AM teams prepare structured customer reviews from conversations Generate an account review or QBR-style summary from recent customer interactions Account ID, optional time period Retrieve all recent calls, emails, and account-linked context. Aggregate themes, requests, blockers, sentiment shifts, and opportunities. Apply a repeatable account review structure. Summarise outcomes, open issues, and suggested next steps. Structured QBR/account review summary with themes, risks, opportunities, and actions Snippet Builder Skill Skill Help users consistently capture and label useful moments from calls Identify useful moments in a call and convert them into reusable snippets Call ID, optional focus area such as pricing, objection, competitor, discovery Retrieve transcript, recording timeline, and call insights. Detect moments matching the requested focus area. Propose snippet start/end times and titles. Tag snippets by type or theme. Optionally save them or add them to a playlist. One or more proposed or saved snippets with title, timing, and tags 💡 Technical details Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Open Questions Track unresolved issues or decisions that need to be made. Update as answers become available. Question Answer Date Answered Competitor research Competitor MCP Availability Core Capabilities Advanced Capabilities Positioning Key Strength Key Weakness Gong ✅ Mature MCP List/search calls Get transcripts get Call metadata Deal intelligence Scorecards, analytics CRM context Enterprise data layer Deep data + ecosystem integrations Limited opinionated AI workflows Avoma ✅ Mature MCP list_meetings get_meeting get_transcript key topics Deal fields/Deal level analysis Coaching/Scoring CRM context pushing insights into Notion / Confluence AI-ready meeting intelligence Strongest “outputs” layer Mostly pull-based interactions Fathom ✅ MCP list/search meetings get transcript get summary/action items Team/org context Webhooks (event-based triggers) get_weekly_recap aggregated action items across meetings creating webhooks directly from the MCP server Lightweight meeting data access Includes summaries + actions Limited analytics depth Grain ✅ Basic MCP get meetings download transcripts Notes Deals Scorecards Data access utility Simple + accessible Low sophistication, sometimes browser-based Outreach ⚠️ MCP (workflow-focused) Search sequences Emails Meetings (Kaia) account / prospect / opportunity lookups Workflow automation Multi-system orchestration Revenue workflow engine End-to-end GTM workflows Less depth in conversation data Attention ⚠️ Emerging MCP Meeting data Transcripts CRM-linked context AI insights Sales assistant workflows create snippets AI sales assistant + CRM Strong CRM linkage Less focus on external knowledge layer Planhat - Related content Ask Jiminny Anything on Call level Galya Dimitrova Hubspot app Galya Dimitrova Product Vision & Strategy Galya Dimitrova Automated CRM Filling Simona Georgieva (Deactivated) Competitive analysis Galya Dimitrova AI agent ideas Steliyan Georgiev Comments James Graham 9 Apr A few areas around Security & Compliance that need more depth before this moves beyond draft: Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate. Team visibility enforcement should be its own HIGH requirement. The current note under access control (\"ensure MCP only exposes data user has access to\") is doing too much work as a single line. Every MCP tool response (especially search_calls and deal-level aggregations) must be filtered through the same visibility engine as the UI. Semantic search is the riskiest here; it could surface results across team boundaries if not properly scoped. No data minimisation or scoping controls. We're exposing transcripts, CRM fields, contact details with job titles, deal context - significant PII and commercial data leaving the platform. Can customers configure which data types are exposed? Can transcripts be PII-stripped before they leave Jiminny? Given we position GDPR consent as a differentiator, we should be thinking about this. GDPR downstream processing is unaddressed. Piping transcripts to OpenAI/Gemini introduces new data processors. Customers need to acknowledge responsibility for downstream processing during setup. We should also consider whether call participants' recording consent covers onward sharing to third-party AI - it almost certainly doesn't by default. This is one for me to pick up. Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users. Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns. Audit logs should go deeper. Current spec (what data, which system, when) is a start. Add: identity behind the key, specific queries/filters used, and failed access attempts (especially cross-visibility-boundary attempts). Enterprise compliance teams will expect this. Downgrade handling. Doc says Scale tier and above, but doesn't say what happens on downgrade. Keys should auto-deactivate. Happy to help draft a dedicated Security & Compliance section if useful. Galya Dimitrova 15 Apr Thanks for the input @James Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.","depth":21,"bounds":{"left":0.41805187,"top":0.19313647,"width":0.051861703,"height":0.11173184},"help_text":"","role_description":"heading","subrole":"AXUnknown"},{"role":"AXStaticText","text":"You said","depth":23,"bounds":{"left":0.36053857,"top":0.19553073,"width":0.019946808,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny MCP Connector","depth":23,"bounds":{"left":0.41805187,"top":0.1963288,"width":0.03125,"height":0.038707104},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Epic","depth":23,"bounds":{"left":0.41805187,"top":0.30806065,"width":0.010305851,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Link to Epik in Jira","depth":23,"bounds":{"left":0.41805187,"top":0.33040702,"width":0.043218084,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Document status","depth":23,"bounds":{"left":0.41805187,"top":0.3527534,"width":0.04105718,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"DRAFT","depth":23,"bounds":{"left":0.41805187,"top":0.37509975,"width":0.016289894,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objective","depth":23,"bounds":{"left":0.41805187,"top":0.39744613,"width":0.024268618,"height":0.016360734},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.","depth":23,"bounds":{"left":0.41805187,"top":0.4197925,"width":0.051529255,"height":0.17278531},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.","depth":23,"bounds":{"left":0.41805187,"top":0.59856343,"width":0.049700797,"height":0.10574621},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👤 Target user","depth":23,"bounds":{"left":0.41805187,"top":0.75458896,"width":0.033909574,"height":0.016759777},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue teams using AI tools (Sales, CS, RevOps)","depth":23,"bounds":{"left":0.41805187,"top":0.7773344,"width":0.051030584,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Companies already experimenting with Claude / OpenAI","depth":23,"bounds":{"left":0.41805187,"top":0.8443735,"width":0.04637633,"height":0.061053474},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)","depth":23,"bounds":{"left":0.41805187,"top":0.9114126,"width":0.051363032,"height":0.0885874},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"🤕 Pain point or problem","depth":23,"bounds":{"left":0.41805187,"top":1.0,"width":0.03706782,"height":-0.0897845},"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny data is locked in the platform","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI tools lack access to high-value conversation context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers must manually export/copy transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Peer pressure as almost all competitors have this","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💥 Impact and benefits","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For customers","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Better AI outputs using real customer conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to build automated workflows across tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increased product stickiness (embedded in workflows)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitive parity with Gong / Avoma etc.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Solution ideas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build an AI Knowledge Connector (MCP-based) that:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Exposes Jiminny data (calls, transcripts, metadata) to AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provides structured insights (summaries, action items, scoring etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Supports deal-level context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Success metrics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Goal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Metric","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"% of customers connecting external AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Number of MCP/API calls per customer - and which ones are being used the most","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Decrease the manual copying of transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"decrease by X %","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User interaction and design","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple setup flow:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate API key / connect MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear instructions for Claude / Zapier setup","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example prompts for Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pre-built workflow templates for Zapier","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Clear documentation - with use cases","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"👉 Focus: usable by non-technical users","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detailed Requirements","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Requirement","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Importance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to access call data through AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny data:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list calls (filter by date, user, account, deal)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript (speaker-separated)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"search calls (keyword + semantic)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support filtering by:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / contact/ lead / deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"jiminny user","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to search and query calls using AI tools","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Support keyword and semantic search across conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide structured insights so customers can generate better outputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each call, expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key points","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Activity type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call Score","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM information","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to understand deal status across multiple calls/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For each deal expose:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"calls/meetings/emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identified risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"contacts related to the deal with their job title","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"linked account details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"all CRM field information that we have in DI + fields used for CRM filling","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example - get_deal_summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"request -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"get_deal_summary\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"responce:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"deal_name\": \"ACME\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"stage\": \"Negotiation\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"recent_activity\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Demo call on 2026-04-01\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing discussion on 2026-04-03\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"],","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"risks\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Pricing concerns\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Security review pending\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"], \"next_steps\": [","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Send follow-up to VP Sales\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"Share security documentation\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"]","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure the AI can easily understand the schema of the data coming from Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Models perform much better when tools have:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"clear names","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rich descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"predictable schemas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide clear, actionable error states for customers when MCP connections fail","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Error messages should guide agents toward solutions with specific suggestions and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to securely control who and what data is accessible externally","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Private meetings shouldn’t be exposed via MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ensure data is structured to be easily combined with:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM (accounts, deals)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"external docs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use consistent identifiers:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal IDs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP is available only for Scale tier and above","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customers who are on Capture tier should be able to access any of their data through MCP.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide visibility into how external AI tools access Jiminny data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Audit logs:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"what data was accessed by which external system when","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"identity - who made the request","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"specific queries/filters used","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"failed access attempts (especially cross-visibility-boundary attempts)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example from Atlassian -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example prompts to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"HIGH","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Expose Jiminny as a Connector in Claude","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Provide customers with example workflows to get started quickly","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to use predefined Jiminny skills for common workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"See examples in table below.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Jiminny UI widget for OpenAI","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"medium","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"The reason to have this is to make certain Jiminny use cases much easier than pure chat.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to run long-running analyses asynchronously and retrieve progress/results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"start the job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let it run in the background","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return a status like “in progress”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"let the client come back for the result later","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"For example - User asks in Claude:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"“Analyse all calls from this quarter and tell me the top 5 objections by segment.”","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"That might require:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"searching lots of calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pulling many summaries/transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregating results","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Instead of making Claude wait 45+ seconds, Jiminny could:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create an analysis job","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"return: job_id = 123, status = running","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"We need this in order to support:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"large-scale analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"multi-call aggregation across lots of data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"playlist generation across many calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"heavy recap/report generation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Creating a snippet through the MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example request:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"tool\": \"create_snippet\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"arguments\": {","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"call_id\": \"call_123\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"start_time\": \"00:18:24\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"end_time\": \"00:19:10\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"example response:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"{","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"snippet_id\": \"snip_987\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"title\": \"Pricing objection\",","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"\"share_url\": \"https://app.jiminny.com/snippets/snip_987\"","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"}","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to add recordings or snippets to a playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Allow AI tools to read reusable Jiminny resources","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"current deal snapshot","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account summary","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"weekly rep recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"approved talk tracks / objection handling guidance","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"curated playlists or snippet libraries","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ability to write coaching notes in Jiminny through MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"low","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"What we want to do for prompts/worflows and skills","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Capability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"User value / job to be done","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Example Jiminny use case","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Inputs","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Steps performed","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Output","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft follow-up email","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps quickly create tailored follow-ups after calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Use the latest customer call to draft an email with recap and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional tone","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional recipient type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve call summary and action items from Jiminny.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull key points, commitments, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply the chosen tone and structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Draft the email in a consistent format.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Customer-ready follow-up email draft","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise call","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users quickly extract the most important points from a conversation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate a concise summary of one call for internal or external use","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional summary style","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional audience","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, metadata, and structured insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify key topics, decisions, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format output for the chosen audience.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Short or structured call summary with actions and risks","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate deal recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers get a fast overview of a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise the current state of a deal based on recent calls and linked CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and meetings linked to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull deal metadata and CRM-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recent changes, blockers, risks, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise into a concise recap.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal recap with status, blockers, risks, and recommended next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare weekly recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prompt","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers or reps review what changed over a time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap of calls, themes, and actions for one rep, team, or account","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"team ID or account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls and summaries for the selected time period.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, action items, and changes.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight notable risks, opportunities, and follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly recap with top themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Prepare deal review","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help reps and managers review deal status using multiple sources of context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Review a deal before pipeline review or forecast meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and meetings tied to the deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull structured call summaries, score, action items, and key points.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Pull linked CRM context such as stage, contacts, and account details.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Format into a review-ready structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal review document with current status, risks, stakeholders, blockers, and next steps","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Weekly team recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers monitor activity and risk across a group of reps or accounts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-of-week team summary for sales leadership","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve calls and structured insights across the team.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Group by rep, account, or deal.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring themes, major risks, strong calls, and missed follow-ups.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Highlight calls or deals that need attention.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team-level recap with patterns, key risks, notable wins, and action areas","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis across calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help teams identify patterns across multiple conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse pricing or competitor objections across recent calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"deal ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"keyword/topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"date range","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search calls matching the topic or keyword.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve summaries, transcripts, and snippets where relevant.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify recurring objections and how they were handled.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Assess whether objections were resolved or remain open.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise patterns and examples.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Objection analysis with themes, frequency, examples, and suggested follow-up actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build onboarding playlist","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help enablement teams curate useful learning content faster","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Build a playlist of strong discovery, demo, or pricing calls for new hires","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Topic","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"call type","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"rep/team filter","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search for calls matching the selected criteria.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rank or filter calls based on score, topics, or key moments.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful snippets or full recordings.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Add selected calls/snippets into a playlist structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate short descriptions for each item.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Curated onboarding playlist with selected calls/snippets and descriptions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal Inspection Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Give users a consistent, opinionated way to assess a deal","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Claude reviews a deal using a repeatable Jiminny methodology","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal ID or account name","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional review scope","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Determine whether the request is suitable for a deal inspection.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a standard Jiminny review framework to assess deal health.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Present findings in a consistent structure with recommended actions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured deal inspection with health assessment, red flags, and recommendations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Manager Coaching Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help managers review calls consistently and identify coaching moments","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Analyse a rep’s call and surface strengths, gaps, and examples for coaching","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional rep ID","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"optional coaching focus","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, score, summary, and key moments from the call.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Evaluate the conversation against a coaching framework.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify what went well and where the rep could improve.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Suggest moments to clip as snippets.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Recommend coaching actions or playlist additions.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CSM / QBR Prep Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help CS or AM teams prepare structured customer reviews from conversations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Generate an account review or QBR-style summary from recent customer interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Account ID, optional time period","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve all recent calls, emails, and account-linked context.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Aggregate themes, requests, blockers, sentiment shifts, and opportunities.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Apply a repeatable account review structure.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Summarise outcomes, open issues, and suggested next steps.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Structured QBR/account review summary with themes, risks, opportunities, and actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Snippet Builder Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Skill","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Help users consistently capture and label useful moments from calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Identify useful moments in a call and convert them into reusable snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Call ID, optional focus area such as pricing, objection, competitor, discovery","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Retrieve transcript, recording timeline, and call insights.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Detect moments matching the requested focus area.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Propose snippet start/end times and titles.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Tag snippets by type or theme.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Optionally save them or add them to a playlist.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"One or more proposed or saved snippets with title, timing, and tags","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"💡 Technical details","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Open Questions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Track unresolved issues or decisions that need to be made. Update as answers become available.","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Question","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Answer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Date Answered","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor research","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Competitor","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"MCP Availability","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Core Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Advanced Capabilities","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Positioning","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Strength","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Key Weakness","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Gong","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"List/search calls","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Get transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get Call metadata","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards, analytics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Enterprise data layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deep data + ecosystem integrations","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited opinionated AI workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Avoma","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Mature MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list_meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_meeting","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"key topics","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deal fields/Deal level analysis","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Coaching/Scoring","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"pushing insights into Notion / Confluence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI-ready meeting intelligence","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strongest “outputs” layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Mostly pull-based interactions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Fathom","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"list/search meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get transcript","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get summary/action items","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Team/org context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Webhooks (event-based triggers)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get_weekly_recap","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"aggregated action items across meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"creating webhooks directly from the MCP server","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Lightweight meeting data access","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Includes summaries + actions","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Limited analytics depth","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Grain","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"✅ Basic MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"get meetings","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"download transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Notes","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Deals","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Scorecards","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Data access utility","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Simple + accessible","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Low sophistication, sometimes browser-based","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Outreach","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ MCP (workflow-focused)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Search sequences","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Emails","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meetings (Kaia)","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"account / prospect / opportunity lookups","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Workflow automation","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Multi-system orchestration","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Revenue workflow engine","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"End-to-end GTM workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less depth in conversation data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Attention","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"⚠️ Emerging MCP","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Meeting data","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Transcripts","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"CRM-linked context","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI insights","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Sales assistant workflows","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"create snippets","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"AI sales assistant + CRM","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Strong CRM linkage","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Less focus on external knowledge layer","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Planhat -","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Related content","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Ask Jiminny Anything on Call level","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Hubspot app","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Product Vision & Strategy","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"},{"role":"AXStaticText","text":"Galya Dimitrova","depth":23,"help_text":"","role_description":"text","subrole":"AXUnknown"}]...
|
5197246770639092577
|
-2099241113185791399
|
idle
|
accessibility
|
NULL
|
Platform Sprint 2 Q2 - Platform Team - Scrum Board Platform Sprint 2 Q2 - Platform Team - Scrum Board - Jira
SevenShores\Hubspot\Exceptions\BadRequest: Client error: `POST [URL_WITH_CREDENTIALS] Graham . I made the product related changes in the table and for the rest I left them as a note in the Technical section for the team to look at when they start looking at this. I wonder about your points related to GDPR. Currently multiple customers are using the Customer API/zapier to feed this data to AI and BI tools. They are also manually exporting transcripts and other information from the platform and feeding it to AI tools. I don’t get how MCP is so much different that all of these and why it will introduce new data processors (which are already being utilised) and would require any changes to the consent provided by participants.
You said
Jiminny MCP Connector
Epic
Link to Epik in Jira
Document status
DRAFT
Objective
Enable customers to connect Jiminny data to external AI tools (Claude, OpenAI, Gemini) so it can be used as part of their broader knowledge base and workflows.
Position Jiminny as a data layer for AI-driven revenue workflows, not just a standalone product.
👤 Target user
Revenue teams using AI tools (Sales, CS, RevOps)
Companies already experimenting with Claude / OpenAI
Mid-market & Enterprise customers with multiple data sources (CRM, docs, CI tools, Support tools etc)
🤕 Pain point or problem
Jiminny data is locked in the platform
AI tools lack access to high-value conversation context
Customers must manually export/copy transcripts
No easy way to combine calls with CRM + docs + other info - using the Jiminny API is not suitable for non technical sales people
Peer pressure as almost all competitors have this
💥 Impact and benefits
For customers
Better AI outputs using real customer conversations
Ability to build automated workflows across tools
For Jiminny
Increased product stickiness (embedded in workflows)
Competitive parity with Gong / Avoma etc.
💡 Solution ideas
Build an AI Knowledge Connector (MCP-based) that:
Exposes Jiminny data (calls, transcripts, metadata) to AI tools
Provides structured insights (summaries, action items, scoring etc.)
Supports deal-level context
Allows integration into customer-owned AI stacks (Claude + Zapier, etc.)
Success metrics
List the key goals and how you will measure success. Include both qualitative and quantitative metrics (e.g., user adoption, retention rate, revenue targets).
Goal
Metric
Increase stickiness by ensuring Jiminny is embedded in the customer’s workflows
% of customers connecting external AI tools
Number of MCP/API calls per customer - and which ones are being used the most
Decrease the manual copying of transcripts
decrease by X %
User interaction and design
Simple setup flow:
Generate API key / connect MCP
Clear instructions for Claude / Zapier setup
Provide:
Example prompts for Claude
Pre-built workflow templates for Zapier
Clear documentation - with use cases
👉 Focus: usable by non-technical users
Detailed Requirements
List the functional and non-functional requirements. Prioritize them (e.g., High, Medium, Low) and add notes for clarification.
Requirement
Importance
Notes
Allow customers to connect Jiminny to their AI tools (e.g. Claude, OpenAI)
HIGH
Provide MCP endpoint so customers can easily integrate Jiminny into their AI stack
Allow customers to access call data through AI tools
HIGH
Expose Jiminny data:
list calls (filter by date, user, account, deal)
get call
get transcript (speaker-separated)
get call metadata
search calls (keyword + semantic)
Support filtering by:
account / contact/ lead / deal
jiminny user
date range
Allow customers to search and query calls using AI tools
HIGH
Support keyword and semantic search across conversations
Provide structured insights so customers can generate better outputs
HIGH
For each call, expose:
Summary
Action items
Key points
Activity type
Notes
Call Score
CRM information
Allow customers to understand deal status across multiple calls/emails
HIGH
For each deal expose:
calls/meetings/emails
Identified risks
contacts related to the deal with their job title
linked account details
all CRM field information that we have in DI + fields used for CRM filling
example - get_deal_summary
request -
{
"tool": "get_deal_summary",
"arguments": {
"deal_name": "ACME"
}
}
responce:
{
"deal_name": "ACME",
"stage": "Negotiation",
"recent_activity": [
"Demo call on 2026-04-01",
"Pricing discussion on 2026-04-03"
],
"risks": [
"Pricing concerns",
"Security review pending"
], "next_steps": [
"Send follow-up to VP Sales",
"Share security documentation"
]
}
Ensure the AI can easily understand the schema of the data coming from Jiminny
HIGH
Models perform much better when tools have:
clear names
rich descriptions
predictable schemas
Provide clear, actionable error states for customers when MCP connections fail
HIGH
Error messages should guide agents toward solutions with specific suggestions and next steps.
Expired or revoked key — “Error: Authentication failed. The API key used to access Jiminny has been revoked or has expired. Please ask your Jiminny admin to generate a new API key in Settings > Organization > General and reconnect the MCP server.”
Rate limit hit — “Error: Rate limit exceeded. Too many requests have been made to the Jiminny MCP server in a short period. Please wait 60 seconds before retrying.”
Permissions mismatch — if a user tries to access data outside their team visibility settings - “Error: Access denied. The requested call (call_id: [X]) is outside your team's visibility settings in Jiminny. Only calls within your permitted scope are accessible.”
Trying to access private meeting - “Error: Access denied. This meeting has been marked as private by the host and cannot be accessed through the MCP connector. Try requesting a different call or ask the meeting host to change the visibility settings in Jiminny.”
Downgraded/Capture account — if a customer's plan drops below Scale tier or the Customer is on Capture in the first place - “Error: MCP access unavailable. The Jiminny MCP connector is available on the Scale plan and above. Your account does not currently have an active Scale subscription.”
Jiminny-side outage — “Error:Jiminny is temporarily unavailable (status 503). This is not a data or permissions issue — please retry in a few minutes.”
No data found - “Error:No calls found matching your criteria. Try adjusting your filters — for example, widening the date range, removing the account filter, or checking the rep name is spelled correctly in Jiminny.”
Allow customers to securely control who and what data is accessible externally
HIGH
Ensure MCP only exposes data user has access to - based on team visibility settings. This means that if a user is not allow to access calls/deals in Jiminny UI then they shouldn’t be able to access those through the MCP as well.
Private meetings shouldn’t be exposed via MCP
HIGH
When a meetings is set to Private then the information about it shouldn’t be exposed through the MCP.
Ensure Jiminny data can be combined with other sources (e.g. CRM, docs) in AI workflows
HIGH
Ensure data is structured to be easily combined with:
CRM (accounts, deals)
external docs
Use consistent identifiers:
account IDs
deal IDs
MCP is available only for Scale tier and above
HIGH
Customers who are on Capture tier should be able to access any of their data through MCP.
This should also cover the scenario when a customer has been on Scale but then downgrades to Capture - their MCP access should stop working.
Provide visibility into how external AI tools access Jiminny data
HIGH
Audit logs:
what data was accessed by which external system when
identity - who made the request
specific queries/filters used
failed access attempts (especially cross-visibility-boundary attempts)
Admins should be able to access the logs from the org settings.We can start with a csv export before building a specific page for it.
example from Atlassian -
Provide customers with example prompts to get started quickly
HIGH
MCP prompts are explicitly defined as prompt templates that clients can discover and call with arguments.
See examples in table below.
Expose Jiminny as a Connector in Claude
medium
Competitors like Fathom are listed as Connectors in Claude. This will make it easier for customers to find us and to connect Jiminny
Provide customers with example workflows to get started quickly
medium
A workflow is usually multiple steps, maybe multiple tools, still focused on one outcome.
See examples in table below.
Allow customers to use predefined Jiminny skills for common workflows
medium
Anthropic describes Skills as folders that include instructions, scripts, and resources that Claude can load when relevant, and as a way to teach Claude repeatable workflows, preferences, and domain expertise once instead of re-explaining them every time. Skills - a reusable package of behavior that can include prompts, instructions, resources, and sometimes scripts/tools for a broader job.
Claude skills - https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf?hsLang=en&utm_source=chatgpt.com
See examples in table below.
Jiminny UI widget for OpenAI
medium
The reason to have this is to make certain Jiminny use cases much easier than pure chat.
Instead of ChatGPT only replying with text like “Here are the last 5 ACME calls,” it could also show a visual Jiminny card or mini-app inside the chat. That widget could list calls, show a deal summary, let the user filter by rep/date/account, open a transcript section, or save a snippet—while the conversation continues in normal chat alongside it.
For Jiminny, the best widget use cases are the ones where users need to browse or compare rather than ask one-off questions. A strong example is deal review: the user asks ChatGPT, “What is the status of the ACME deal?” ChatGPT calls your Jiminny MCP tools, and the widget shows a compact deal view with recent calls, risks, next steps, stakeholders, and links to transcripts/snippets. The chat can still summarize the answer in words, but the widget gives a clearer, scannable interface.
Another good Jiminny widget use case is call search and exploration. A text answer can say “I found 12 calls mentioning pricing objections,” but a widget could show a sortable list with date, rep, account, score, and objection count, then let the user click into one call to see the exact excerpt or jump to a transcript section.
Allow customers to run long-running analyses asynchronously and retrieve progress/results
low
Some requests/questions require analysis of Hugh volumes of data. For those, it may be better to:
start the job
let it run in the background
return a status like “in progress”
let the client come back for the result later
For example - User asks in Claude:
“Analyse all calls from this quarter and tell me the top 5 objections by segment.”
That might require:
searching lots of calls
pulling many summaries/transcripts
aggregating results
Instead of making Claude wait 45+ seconds, Jiminny could:
create an analysis job
return: job_id = 123, status = running
We need this in order to support:
large-scale analysis
multi-call aggregation across lots of data
playlist generation across many calls
heavy recap/report generation
Creating a snippet through the MCP
low
Create a shareable snippet from a call recording by specifying a time range. Useful for highlighting key moments — objection handling, pricing discussions, competitive mentions — and sharing them with your team.
example request:
{
"tool": "create_snippet",
"arguments": {
"call_id": "call_123",
"start_time": "00:18:24",
"end_time": "00:19:10",
"title": "Pricing objection"
}
}
example response:
{
"snippet_id": "snip_987",
"title": "Pricing objection",
"share_url": "https://app.jiminny.com/snippets/snip_987"
}
Ability to add recordings or snippets to a playlist
low
Allow customers to create libraries of calls automatically. They can use those for onboarding or other purposes
Allow AI tools to read reusable Jiminny resources
low
fathom
Resources are useful for data that should be read consistently rather than recomputed every time. In Jiminny terms, that could be:
current deal snapshot
account summary
weekly rep recap
approved talk tracks / objection handling guidance
curated playlists or snippet libraries
This is useful because an AI can consume these directly as context, instead of repeatedly stitching together many tool calls.
Ability to write coaching notes in Jiminny through MCP
low
What we want to do for prompts/worflows and skills
Capability
Type
User value / job to be done
Example Jiminny use case
Inputs
Steps performed
Output
Draft follow-up email
Prompt
Help reps quickly create tailored follow-ups after calls
Use the latest customer call to draft an email with recap and next steps
Call ID
optional tone
optional recipient type
Retrieve call summary and action items from Jiminny.
Pull key points, commitments, and next steps.
Apply the chosen tone and structure.
Draft the email in a consistent format.
Customer-ready follow-up email draft
Summarise call
Prompt
Help users quickly extract the most important points from a conversation
Generate a concise summary of one call for internal or external use
Call ID
optional summary style
optional audience
Retrieve transcript, metadata, and structured insights.
Identify key topics, decisions, risks, and next steps.
Format output for the chosen audience.
Short or structured call summary with actions and risks
Generate deal recap
Prompt
Help reps and managers get a fast overview of a deal
Summarise the current state of a deal based on recent calls and linked CRM context
Deal ID or account name
optional date range
Retrieve calls and meetings linked to the deal.
Pull deal metadata and CRM-linked context.
Identify recent changes, blockers, risks, and next steps.
Summarise into a concise recap.
Deal recap with status, blockers, risks, and recommended next steps
Prepare weekly recap
Prompt
Help managers or reps review what changed over a time period
Weekly recap of calls, themes, and actions for one rep, team, or account
Rep ID
team ID or account ID
time period
Retrieve relevant calls and summaries for the selected time period.
Aggregate themes, action items, and changes.
Highlight notable risks, opportunities, and follow-ups.
Weekly recap with top themes, risks, opportunities, and actions
Prepare deal review
Workflow
Help reps and managers review deal status using multiple sources of context
Review a deal before pipeline review or forecast meeting
Deal ID
optional date range
Retrieve all recent calls, emails, and meetings tied to the deal.
Pull structured call summaries, score, action items, and key points.
Pull linked CRM context such as stage, contacts, and account details.
Aggregate open risks, unresolved objections, stakeholder gaps, and next steps.
Format into a review-ready structure.
Deal review document with current status, risks, stakeholders, blockers, and next steps
Weekly team recap
Workflow
Help managers monitor activity and risk across a group of reps or accounts
End-of-week team summary for sales leadership
Team ID
date range
Retrieve calls and structured insights across the team.
Group by rep, account, or deal.
Identify recurring themes, major risks, strong calls, and missed follow-ups.
Highlight calls or deals that need attention.
Team-level recap with patterns, key risks, notable wins, and action areas
Objection analysis across calls
Workflow
Help teams identify patterns across multiple conversations
Analyse pricing or competitor objections across recent calls
Account ID
deal ID
rep ID
keyword/topic
date range
Search calls matching the topic or keyword.
Retrieve summaries, transcripts, and snippets where relevant.
Identify recurring objections and how they were handled.
Assess whether objections were resolved or remain open.
Summarise patterns and examples.
Objection analysis with themes, frequency, examples, and suggested follow-up actions
Build onboarding playlist
Workflow
Help enablement teams curate useful learning content faster
Build a playlist of strong discovery, demo, or pricing calls for new hires
Topic
call type
rep/team filter
optional time period
Search for calls matching the selected criteria.
Rank or filter calls based on score, topics, or key moments.
Identify useful snippets or full recordings.
Add selected calls/snippets into a playlist structure.
Generate short descriptions for each item.
Curated onboarding playlist with selected calls/snippets and descriptions
Deal Inspection Skill
Skill
Give users a consistent, opinionated way to assess a deal
Claude reviews a deal using a repeatable Jiminny methodology
Deal ID or account name
optional review scope
Determine whether the request is suitable for a deal inspection.
Retrieve relevant calls, summaries, CRM context, stakeholders, and risks.
Apply a standard Jiminny review framework to assess deal health.
Look for missing next steps, unresolved objections, stakeholder gaps, and warning signs.
Present findings in a consistent structure with recommended actions.
Structured deal inspection with health assessment, red flags, and recommendations
Manager Coaching Skill
Skill
Help managers review calls consistently and identify coaching moments
Analyse a rep’s call and surface strengths, gaps, and examples for coaching
Call ID
optional rep ID
optional coaching focus
Retrieve transcript, score, summary, and key moments from the call.
Evaluate the conversation against a coaching framework.
Identify what went well and where the rep could improve.
Suggest moments to clip as snippets.
Recommend coaching actions or playlist additions.
Coaching review with strengths, gaps, snippet suggestions, and recommended coaching actions
CSM / QBR Prep Skill
Skill
Help CS or AM teams prepare structured customer reviews from conversations
Generate an account review or QBR-style summary from recent customer interactions
Account ID, optional time period
Retrieve all recent calls, emails, and account-linked context.
Aggregate themes, requests, blockers, sentiment shifts, and opportunities.
Apply a repeatable account review structure.
Summarise outcomes, open issues, and suggested next steps.
Structured QBR/account review summary with themes, risks, opportunities, and actions
Snippet Builder Skill
Skill
Help users consistently capture and label useful moments from calls
Identify useful moments in a call and convert them into reusable snippets
Call ID, optional focus area such as pricing, objection, competitor, discovery
Retrieve transcript, recording timeline, and call insights.
Detect moments matching the requested focus area.
Propose snippet start/end times and titles.
Tag snippets by type or theme.
Optionally save them or add them to a playlist.
One or more proposed or saved snippets with title, timing, and tags
💡 Technical details
Auth model needs specifying. Are API keys scoped per user, per team, or per Jiminny instance? If a key is generated by an admin, does it inherit their visibility level, or provide org-wide? Are permissions evaluated at query time or frozen at key creation? Needs to be query-time, otherwise permission changes won't propagate.
Key lifecycle management is missing. No mention of rotation, expiry, or revocation. A departed employee's active key is a data leak. Need: revocation, optional expiry, alerts on stale keys, and consider whether keys should be tied to service accounts rather than personal users.
Rate limiting should be HIGH. A compromised key or misconfigured automation could bulk-export an entire customer's call history. Need per-key rate limits and anomaly detection on unusual access patterns.
Open Questions
Track unresolved issues or decisions that need to be made. Update as answers become available.
Question
Answer
Date Answered
Competitor research
Competitor
MCP Availability
Core Capabilities
Advanced Capabilities
Positioning
Key Strength
Key Weakness
Gong
✅ Mature MCP
List/search calls
Get transcripts
get Call metadata
Deal intelligence
Scorecards, analytics
CRM context
Enterprise data layer
Deep data + ecosystem integrations
Limited opinionated AI workflows
Avoma
✅ Mature MCP
list_meetings
get_meeting
get_transcript
key topics
Deal fields/Deal level analysis
Coaching/Scoring
CRM context
pushing insights into Notion / Confluence
AI-ready meeting intelligence
Strongest “outputs” layer
Mostly pull-based interactions
Fathom
✅ MCP
list/search meetings
get transcript
get summary/action items
Team/org context
Webhooks (event-based triggers)
get_weekly_recap
aggregated action items across meetings
creating webhooks directly from the MCP server
Lightweight meeting data access
Includes summaries + actions
Limited analytics depth
Grain
✅ Basic MCP
get meetings
download transcripts
Notes
Deals
Scorecards
Data access utility
Simple + accessible
Low sophistication, sometimes browser-based
Outreach
⚠️ MCP (workflow-focused)
Search sequences
Emails
Meetings (Kaia)
account / prospect / opportunity lookups
Workflow automation
Multi-system orchestration
Revenue workflow engine
End-to-end GTM workflows
Less depth in conversation data
Attention
⚠️ Emerging MCP
Meeting data
Transcripts
CRM-linked context
AI insights
Sales assistant workflows
create snippets
AI sales assistant + CRM
Strong CRM linkage
Less focus on external knowledge layer
Planhat -
Related content
Ask Jiminny Anything on Call level
Galya Dimitrova
Hubspot app
Galya Dimitrova
Product Vision & Strategy
Galya Dimitrova...
|
NULL
|