- Home
- Skills
- Business & Operations
- Meeting Notes: Decisions & Action Items
Works with the AI Agents you already use
Meeting Notes: Decisions & Action Items
Turns transcripts or messy notes into minutes: decisions with rationale, action items with owner and due date, open questions, risks, a running DECISIONS.md, a follow-up email draft and tasks for Jira, Linear or Asana.
Free
Meeting Notes: Decisions & Action Items
Example session with this skill installed
Summarize this meeting transcript: Sarah says we go with React for the frontend because of the talent pool, even though Dave suggested Vue. Dave to set up the repo by Friday. Who pays for AWS?
- Read your context and instructions
- Compiled the meeting notes
I have analyzed the transcript and updated your project documentation. I have added the React decision to DECISIONS.md and prepared the follow-up tasks.
Meeting Minutes - 2025-10-14
Attendees: Sarah, Dave
TL;DR
The team decided to use React for the frontend development to simplify hiring. Dave is responsible for the initial repository setup by the end of the week. Budget allocation for AWS remains an open question.
Decisions
| # | Decision | Rationale | Decided by | Status |
|---|---|---|---|---|
| 1 | Use React for frontend | Larger talent pool for hiring | Sarah | Firm |
Action items
| # | Owner | Action | Due | Depends on |
|---|---|---|---|---|
| 1 | Dave | Set up project repository | 2025-10-17 | - |
Open questions
| Question | Owner | Needed by |
|---|---|---|
| Who is responsible for the AWS costs? | Not set (?) | Next Sync |
DECISIONS.md Entry
- 2025-10-14 · Use React for frontend · why: larger talent pool for hiring; rejected Vue · by: Sarah · source: Frontend Sync · supersedes: -
Tasks Export (CSV)
title,owner,due,description
Set up project repository,Dave,2025-10-17,Initialize the React frontend repository.
Next steps
- Review the
DECISIONS.mdfile to ensure the rationale is captured correctly. - Copy the CSV block above to import the task into your project management tool.
- Send the drafted summary to Sarah and Dave for confirmation on the AWS question.
Connects securely to your tools. The creator never sees your data.
About this skill
Meeting transcripts often end up as dead files where decisions are buried and action items are forgotten. This skill parses messy notes and raw transcripts to extract a structured record of truth. It ensures every decision has a rationale and every task has a clear owner.
What it does
- Extract decisions identifies agreed-upon choices along with their rationale and any rejected alternatives.
- Assign action items maps specific verbs and objects to individual owners with due dates and dependencies.
- Track open questions logs unresolved issues and assigns them to the relevant parties for follow-up.
- Maintain DECISIONS.md appends new firm decisions to a persistent log to prevent contradictory choices.
- Generate follow-ups drafts email or Slack summaries and prepares task exports for project management tools.
How it works
- Input processing scans the provided transcript or chat thread to identify attendees and dates.
- Analysis phase reads the entire context to distinguish between suggestions and firm agreements.
- Minutes generation creates a structured summary including a 5-line TL;DR and formatted tables.
- Log persistence updates a running
DECISIONS.mdfile, checking for decisions that supersede previous ones.
Frameworks & tools
This skill works with raw text exports from Zoom, Teams, Google Meet, or Otter. It outputs Markdown, CSV for Jira/Linear/Asana imports, and formatted email drafts.
Why this beats prompting it yourself
Generic prompts often hallucinate task owners or miss the "why" behind a decision. This skill enforces a strict data schema that tracks rationale and dependencies, ensuring your project history is searchable and audit-ready.
Use cases
- Post-meeting debriefs converting a 60-minute Zoom transcript into a one-page summary.
- Project auditing maintaining a
DECISIONS.mdfile to track why specific technical paths were chosen. - Task synchronization generating CSV exports to bulk-upload action items into Linear or Jira.
- Stakeholder updates drafting neutral, professional follow-up emails for clients or leadership.
Known limitations
Does not support audio files directly, requires a text transcript. Will not invent owners or dates if they are missing from the source text.
Requires text transcripts or notes (no direct audio). Does not support auto-assigning tasks to people not mentioned in the transcript. Use of speaker labels is required for owner attribution.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
This skill has been used 7 times.
Trust & safety
Security scan
Security scan passed 7 days ago
- Free with an Agensi account