|
72203
|
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...
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72203
|
|
72204
|
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...
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72204
|
|
72205
|
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 ac...
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72205
|
|
72206
|
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)...
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72206
|
|
72207
|
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)...
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72207
|
|
72208
|
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 ac...
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72208
|
|
72209
|
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 ac...
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72209
|
|
72210
|
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...
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72210
|
|
72211
|
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 ac...
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72211
|
|
72212
|
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 ac...
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72212
|
|
72213
|
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....
|
Firefox
|
Architecture overview - Model Context Protocol — W Architecture overview - Model Context Protocol — Work...
|
modelcontextprotocol.io/docs/learn/architecture
|
72213
|
|
15467
|
DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Weste DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 6TB (WD60EFPX)
Western Digital Red Plus 6TB (WD60EFPX)
Количка
Количка
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
2.3
KB/s
919
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
15467
|
|
15472
|
DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Weste DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 6TB (WD60EFPX)
Western Digital Red Plus 6TB (WD60EFPX)
Количка
Количка
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
8.7
KB/s
14.5
KB/s...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
15472
|
|
15473
|
DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Weste DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 6TB (WD60EFPX)
Western Digital Red Plus 6TB (WD60EFPX)
Количка
Количка
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
8.7
KB/s
14.5
KB/s...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
15473
|
|
15521
|
Task Manager
DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] Task Manager
DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 6TB (WD60EFPX)
Western Digital Red Plus 6TB (WD60EFPX)
Количка
Количка
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
1.5
KB/s
324
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
15521
|
|
15555
|
DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Weste DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 6TB (WD60EFPX)
Western Digital Red Plus 6TB (WD60EFPX)
Количка
Количка
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
0
B/s
0
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
15555
|
|
15556
|
DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Weste DXP4800PLUS-B5F8
Inbox (7) - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 6TB (WD60EFPX)
Western Digital Red Plus 6TB (WD60EFPX)
Количка
Количка
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
0
B/s
0
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
15556
|
|
23136
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
Jiminny x Shiji - Reconnecting the platform
Jiminny x Shiji - Reconnecting the platform
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
7.7
KB/s
697
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker
-
Folder
2026-03-30 19:14
BT
-
Folder
2026-01-13 12:37
bookdrop
-
Folder
2025-12-26 18:16
Backup
-
Folder
2026-04-06 19:44
Audiobooks
-
Folder
2026-02-10 10:22
Total items: 19
Control Panel
Search
Connection & Access
User Management
File Service
Device Connection
Domain/LDAP
Terminal
General
Hardware & Power
Time & Language
Network
Security
Indexing Service
Service
About
Update & Restore
Telnet
Enable
Enable
Port
23
Advanced settings
SSH
Enable
Enable
Port
22
Shut down automatically
30 min later
2026-04-14 20:59 will automatically shut down
Advanced settings
Function description
Use a terminal to log in and manage your system. When enabling this function, it is recommended to set a strong password for the login account and enable
auto block
auto block
to enhance system security.
Apply
Files
Personal Folder
Shared Folder
Youtube
Work
Test
Škola
SyncData
screenpipe
pipes
data
Jazyky
Gitara
Computer Science
CODE
screenpipe
Music
Movies
Media
Marti
Google
games
Family tree documents
ebooks
Documents
docker
BT
bookdrop
Backup
Audiobooks
Apps
User Folder
Tag
Recycle Bin
...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
23136
|
|
23137
|
Jiminny x Shiji - Reconnecting the platform
app.ji Jiminny x Shiji - Reconnecting the platform
app.jiminny.com
DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
Jiminny x Shiji - Reconnecting the platform
Jiminny x Shiji - Reconnecting the platform
Close tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
7.7
KB/s
697
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker
-
Folder
2026-03-30 19:14
BT
-
Folder
2026-01-13 12:37
bookdrop
-
Folder
2025-12-26 18:16
Backup
-...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
23137
|
|
23667
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Close tab
| Senetic
Close tab
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Close tab
SQLite Web: db.sqlite
Close tab
Screenpipe Dashboard
Close tab
Welcome to Steam
Close tab
New Tab
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
Customize sidebar
8.1
KB/s
16.2
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker
-
Folder
2026-03-30 19:14
BT
-
Folder
2026-01-13 12:37
bookdrop
-...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
23667
|
|
23670
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Close tab
| Senetic
Close tab
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Close tab
SQLite Web: db.sqlite
Close tab
Screenpipe Dashboard
Close tab
Welcome to Steam
Close tab
New Tab
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
Customize sidebar
8.1
KB/s
16.2
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker
-
Folder
2026-03-30 19:14
BT
-
Folder
2026-01-13 12:37
bookdrop
-
Folder
2025-12-26 18:16
Backup
-
Folder
2026-04-06 19:44
Audiobooks
-
Folder
2026-02-10 10:22
Total items: 19
Control Panel
Search
Connection & Access
User Management
File Service
Device Connection
Domain/LDAP
Terminal
General
Hardware & Power
Time & Language
Network
Security
Indexing Service
Service
About
Update & Restore
Telnet
Enable
Enable
Port
23
Advanced settings
SSH
Enable
Enable
Port
22
Shut down automatically
30 min later
2026-04-14 20:59 will automatically shut down
Advanced settings
Function description
Use a terminal to log in and manage your system. When enabling this function, it is recommended to set a strong password for the login account and enable
auto block
auto block
to enhance system security.
Apply
Files
Personal Folder
Shared Folder
Youtube
Work
Test
Škola
SyncData
screenpipe
pipes
data
Jazyky
Gitara
Computer Science
CODE
screenpipe
Music
Movies
Media
Marti
Google
games
Family tree documents
ebooks
Documents
docker
BT
bookdrop
Backup
Audiobooks
Apps
User Folder
Tag
Recycle Bin
Shared Folder
Test...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
23670
|
|
23673
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Close tab
| Senetic
Close tab
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Close tab
SQLite Web: db.sqlite
Close tab
Screenpipe Dashboard
Close tab
Welcome to Steam
Close tab
New Tab
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
Customize sidebar
4.7
KB/s
2.5
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker
-
Folder
2026-03-30 19:14
BT
-
Folder
2026-01-13 12:37
bookdrop
-
Folder
2025-12-26 18:16
Backup
-
Folder
2026-04-06 19:44
Audiobooks
-
Folder
2026-02-10 10:22
Total items: 19
Control Panel
Search
Connection & Access
User Management
File Service
Device Connection
Domain/LDAP
Terminal
General
Hardware & Power
Time & Language
Network
Security
Indexing Service
Service
About
Update & Restore
Telnet
Enable
Enable
Port
23
Advanced settings...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
23673
|
|
23675
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
4.7
KB/s
2.5
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker
-
Folder
2026-03-30 19:14
BT
-
Folder...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
23675
|
|
23687
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
1.9
KB/s
4.8
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker
-
Folder
2026-03-30 19:14
BT
-
Folder
2026-01-13 12:37
bookdrop
-
Folder
2025-12-26 18:16
Backup
-
Folder
2026-04-06 19:44
Audiobooks
-
Folder
2026-02-10 10:22
Total items: 19
Control Panel
Search
Connection & Access
User Management
File Service
Device Connection
Domain/LDAP
Terminal
General
Hardware & Power
Time & Language
Network
Security
Indexing Service
Service
About
Update & Restore
Telnet
Enable
Enable
Port
23
Advanced settings
SSH
Enable
Enable
Port
22
Shut down automatically
3h later
2026-04-15 17:29 will automatically shut down
Advanced settings
Function description
Use a terminal to log in and manage your system. When enabling this function, it is recommended to set a strong password for the login account and enable
auto block
auto block
to enhance system security.
Apply
Files
Personal Folder
Shared Folder
Youtube
Work
...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
23687
|
|
23688
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
2.3
KB/s
739
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker
-
Folder
2026-03-30 19:14
BT
-
Folder
2026-01-13 12:37
bookdrop
-
Folder
2025-12-26 18:16
Backup
-
Folder
2026-04-06 19:44
Audiobooks
-
Folder
2026-02-10 10:22
Total items: 19
Control Panel
Search
Connection & Access
User Management
File Service
Device Connection
Domain/LDAP
Terminal
General
Hardware & Power
Time & Language
Network
Security
Indexing Service
Service
About
Update & Restore...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
23688
|
|
23690
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
818
B/s
192
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker
-
Folder
2026-03-30 19:14
BT
-
Folder
2026-01-13 12:37
bookdrop
-
Folder
2025-12-26 18:16
Backup
-
Folder
2026-04-06 19:44
Audiobooks
-
Folder
2026-02-10 10:22
Total items: 19
Control Panel
Search
Connection & Access
User Management
File Service
Device Connection
Domain/LDAP
Terminal
General
Hardware & Power
Time & Language
Network
Security
Indexing Service
Service
About
Update & Restore
Telnet
Enable
Enable
Port
23
Advanced settings...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
23690
|
|
23692
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
0
B/s
0
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Files
Personal Folder
Shared Folder
User Folder
Tag
Recycle Bin
Shared Folder
Shared Folder
Please enter
Name
Size
Type
Modification date
Youtube
-
Folder
2026-01-15 17:57
Work
-
Folder
2026-03-13 10:17
Test
-
Folder
2026-04-11 15:46
screenpipe
-
Folder
2026-04-13 19:14
Music
-
Folder
2025-10-24 22:41
Movies
-
Folder
2026-02-09 20:06
Media
-
Folder
2025-09-27 17:56
Marti
-
Folder
2025-10-24 20:06
Google
-
Folder
2025-11-21 13:20
games
-
Folder
2026-01-01 15:39
Family tree documents
-
Folder
2026-01-25 17:51
ebooks
-
Folder
2026-03-16 10:34
Documents
-
Folder
2025-10-14 12:21
docker
-
Folder
2026-03-30 19:14...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
23692
|
|
46018
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Close Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
AI Chat settings
Close
Google Account: Lukáš Koválik ([EMAIL])
Main menu
New chat
Gemini
Temporary chat
PLUS
PLUS
Conversation with Gemini
Conversation with Gemini
Hi Lukáš
Where should we start?
Where should we start?
🖼️ Create image, button, tap to use tool
🖼️ Create image
🎸 Create music, button, tap to use tool
🎸 Create music
Boost my day, button, tap to use tool
Boost my day
Help me learn, button, tap to use tool
Help me learn
Write anything, button, tap to use tool
Write anything
Create a video, button, tap to use tool
Create a video
Dismiss
Sharper images, better text. Meet Nano Banana 2.
Try it
Try it
Ask Gemini
Ask Gemini
Open upload file menu
Tools...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46018
|
|
46024
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
12.3
KB/s
6.7
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46024
|
|
46032
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
10.5
KB/s
1.2
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46032
|
|
46095
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
27.4
KB/s
2.1
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46095
|
|
46119
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
8.5
KB/s
4.7
KB/s
Files
Control Panel
Storage...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46119
|
|
46120
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
8.5
KB/s
4.7
KB/s
Files
Control Panel...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46120
|
|
46123
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
6.5
KB/s
14.5
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46123
|
|
46127
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
25.4
KB/s
15.8
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46127
|
|
46128
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
752
B/s
132
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46128
|
|
46562
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
9.1
KB/s
832
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46562
|
|
46563
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
9.1
KB/s
832
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46563
|
|
46570
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
7.7
KB/s
14.6
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46570
|
|
46579
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
4.3
KB/s
10.5
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46579
|
|
46594
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
2
KB/s
659
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
46594
|
|
47276
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
3.6
KB/s
851
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
47276
|
|
47280
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
47280
|
|
47281
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
47281
|
|
47598
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
47598
|
|
47599
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab...
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
47599
|
|
47610
|
SlackFileEoitViewHistoryWindowHelpnds.lakylak.xyc/ SlackFileEoitViewHistoryWindowHelpnds.lakylak.xyc/aesklootp! Western Digital Red Plus 3.5 6TB 5400rpm 256MB SAS I SeneticТвърд диск, Western Digital Red 6TB Plus ( 3.5", 256N@ SQLite Web: db.sqlitew Screenpipe DashboardV Welcome to Steam• YoulubeNew Tab— New TabFilesmneelercontrol PanelPhotosStorageOnline OfficeApp centereXI=OILogsVirtual MachineSupportlesk ManceeDownloadsDLNAuniversal bearenFlle version explorerMusicSecurityVoue DivesJellvfin-HT• Suoport Dailv • in 8m100% zFri 17 Aor 14:52:25од пай вековSAN ManadervaultSnapshoVOmICSSync & BackupGo to mark......
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
47610
|
|
47616
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
1.9
KB/s
2.4
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Go to mark......
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
47616
|
|
47617
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
2.6
KB/s
4.4
KB/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Go to mark......
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
47617
|
|
49240
|
DXP4800PLUS-B5F8
Steam Account Verification - [EMA DXP4800PLUS-B5F8
Steam Account Verification - [EMAIL] - Gmail
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
Western Digital Red Plus 3.5 6TB 5400rpm 256MB SATA3 (WD60EFPX) от 238,97 € (467,38 лв.) Вътрешен хард диск Western Digital - Pazaruvaj.com
| Senetic
| Senetic
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
Твърд диск, Western Digital Red 6TB Plus ( 3.5", 256MB, 5400
SQLite Web: db.sqlite
SQLite Web: db.sqlite
Screenpipe Dashboard
Screenpipe Dashboard
Welcome to Steam
Welcome to Steam
YouTube
YouTube
New Tab
New Tab
New Tab
Customize sidebar
Open Google Gemini (⌃X)
Open history (⇧⌘H)
Open bookmarks (⌘B)
Bitwarden
6.5
KB/s
738
B/s
Files
Control Panel
Storage
App Center
Logs
Support
Task Manager
Universal Search
Music
Cloud Drives
Theater
Photos
Online Office
TextEdit
Virtual Machine
Downloads
DLNA
File Version Explorer
Security
Jellyfin-HT
SAN Manager
Vault
Snapshot
Comics
Sync & Backup
Go to mark......
|
Firefox
|
DXP4800PLUS-B5F8 — Personal
|
nas.lakylak.xyz/desktop/#/
|
49240
|