Meeting Memory
A folder structure for the meetings that matter. Decisions, follow-ups, attendees. Claude and ChatGPT read it the next time anyone asks "what did we decide last week".
Requires an AI connected to your Hjarni account via MCP.
Copy this URL and paste it into Claude or ChatGPT to install the template.
How to use
- 1 Share this page. Paste this URL into Claude or ChatGPT. Your AI reads the template definition and installs it.
- 2 Folders, tags, and instructions appear. Your AI creates the full structure in your Hjarni account, ready to use.
- 3 Start adding notes. The AI instructions guide your AI on where to put things and how to organize them.
A folder structure for the meetings that matter. Claude and ChatGPT read it the next time anyone asks "what did we decide last week".
Meeting notes die fast. They get saved in five different places. Six months later, no one remembers why the team chose option B, or who agreed to follow up. This template gives those notes a shape your AI can actually use.
A specific note format
Every meeting note follows the same shape so the AI knows where to look. The starter note in each meeting folder is this skeleton.
# Weekly Product Sync - 2026-05-16
Attendees:
Decisions:
Follow-ups:
Open questions:
Context for next time:
Follow-ups, not task management
Follow-ups are notes, not a task system. Use them to remember what was agreed and who said they would do it. Keep execution wherever your team already works: Linear, GitHub Issues, Notion, a notebook. Hjarni is the memory, not the project manager.
The extraction workflow
- Save the raw meeting note in the right folder.
- Ask Claude or ChatGPT to extract decisions and follow-ups.
- Save the extracted notes into the Decisions and Follow-ups folders, cross-linking back to the source meeting.
This keeps raw notes faithful and the synthesis searchable.
Before the next meeting
Ten minutes before your next weekly sync, ask Claude or ChatGPT: "Brief me on the last three product syncs and list unresolved decisions." Claude reads the recurring meeting folder and gives you the thread, not just the latest note. You walk in already caught up.
A workflow that earns the template's keep
After a meeting, paste the notes into the right folder. The next time you ask "what did we agree on Tuesday", your AI reads the note and tells you. The next time someone new joins the team, they connect their Claude and ask "what are our open decisions" and get a real answer.
The same notes serve two readers. You write them once. You and your AI both remember.
Common questions
Common questions
Does this transcribe meetings?
No. Hjarni stores the notes and context your AI can read. Bring notes from your meeting tool, transcript tool, or manual notes.
Is this a replacement for Asana or Linear?
No. Follow-ups are searchable memory of what was agreed. Keep your task system where it is.
Can my team share this?
Yes. Install in a team space so every teammate's Claude or ChatGPT reads from the same meetings. See Hjarni for Teams.
Does the AI run my meetings?
No. The AI reads the notes you save after the fact.
Related pages
Structure
Tags
Folders
For your AI
Share this page with your AI. It reads the definition below, creates the folders, instructions, tags, and starter notes in your account.
Show template definition
Install steps for AI agents:
- Check existing tags with
tags-list. Only create missing ones withtags-create. - Create containers top-down using
containers-create, noting the returned IDs. Useparent_idto build the hierarchy. - For each container with
llm_instructions, callinstructions-updatewithlevel: "container"and the container's ID. - Create any seed notes using
notes-create, placing them in the correct container by ID. Usecontainer_pathto resolve which container. - Discuss any customizations with the user before or after installing.
---
name: Meeting Memory
description: 'A folder structure for the meetings that matter. Decisions, follow-ups,
attendees. Claude and ChatGPT read it the next time anyone asks "what did we decide
last week".
'
tags:
- decision
- follow-up
- blocker
- context-for-next-time
containers:
- name: Meeting Memory
description: The shared memory of every meeting that mattered. Searchable, not a
task system.
llm_instructions: |
This is meeting memory. Notes stay faithful to what was said. Synthesis lives in dedicated folders.
- Never invent attendees, decisions, or follow-ups. If the raw note does not contain a detail, omit it.
- When asked "what did we decide about X" or "what did we agree on Tuesday", search Decisions and Follow-ups first, then the source meeting note.
- Always cite the source meeting by title and date when quoting from a meeting note.
- Follow-ups are notes, not a live task system. Do not pretend they are tracked or assigned in an external tool.
- Suggest filing extracted decisions and follow-ups back into the right folder after each meeting.
children:
- name: Recurring Meetings
description: One folder per recurring meeting. Dated notes inside.
llm_instructions: |
Use this folder for the meetings that happen every week or month.
- Each recurring meeting has its own sub-folder when it earns one (weekly standup, monthly review).
- Notes follow the shipped skeleton: Attendees, Decisions, Follow-ups, Open questions, Context for next time.
- When asked about a recurring meeting, return the most recent note first, then summarize the trend across the last four.
- Tag every note with "meeting".
- name: One-off Meetings
description: Project kickoffs, customer calls, interviews.
llm_instructions: |
Use this folder for meetings that do not repeat.
- One note per meeting. Title format: "<Meeting topic> - YYYY-MM-DD".
- Use the same shipped skeleton as recurring meetings.
- Cross-link related notes (related projects, related people) using wiki-links.
- name: Decisions
description: Extracted decisions, cross-linked back to the meeting note that produced
them.
llm_instructions: |
Use this folder for the decisions extracted from meetings.
- One decision per note. Title is a short noun phrase.
- Always cite the source meeting and date. Quote the decision exactly from the source.
- Structure each note with: Context, Decision, Tradeoffs accepted, Source meeting.
- When a new decision supersedes an old one, create a new note and link the two.
- Tag every note with "decision".
- name: Follow-ups
description: Searchable memory of what was agreed. Not a task system, see the
template page.
llm_instructions: |
Use this folder for follow-ups extracted from meetings.
- One follow-up per note. Title format: "<Owner>: <Short task>".
- When asked what was agreed, return the matching follow-ups with their source meeting.
- These are notes, not a live task list. Do not pretend they are tracked in Linear, Asana, or GitHub Issues.
- When the user moves a follow-up to an external task system, suggest tagging the note "resolved" or moving it to an archive.
- Tag every note with "follow-up".
notes:
- title: Weekly Product Sync - 2026-05-16
body: |
A starter note showing the meeting skeleton. Replace it with your real notes.
**Attendees:** Sarah, Marcos, Jess
**Decisions:**
- Push launch from April to May.
- Drop the second onboarding email until volume justifies it.
**Follow-ups:**
- Sarah: update the public roadmap by Friday.
- Marcos: confirm with three pilot customers that the new date works.
**Open questions:**
- Do we publish a delay note on the blog or wait for the new launch post?
**Context for next time:**
- Pilot customers may need a heads-up before the public delay note.
This is a starter note. Replace it with a real meeting note.
tags:
- decision
- follow-up
container_path: Meeting Memory > Recurring Meetings
- title: 'Customer call: Acme Corp - 2026-05-16'
body: |
A starter note for a one-off meeting. Replace it with a real call note.
**Attendees:** Marcos, three Acme product leads
**Decisions:**
- Acme commits to a paid pilot starting June.
**Follow-ups:**
- Marcos: send the pilot agreement by Monday.
**Open questions:**
- Do they need SSO before the pilot starts, or only after?
**Context for next time:**
- They asked twice about an audit log. Worth raising in the next product sync.
This is a starter note. Replace it with a real customer call note.
tags:
- follow-up
container_path: Meeting Memory > One-off Meetings
- title: 'Decision: Push launch to May'
body: |
## Context
April timing put the launch on top of two competitors' announcements.
## Decision
Move the launch date from April to May.
## Tradeoffs accepted
- One month of additional runway burn.
- Risk of waning team momentum, offset by extra polish.
## Source meeting
[[Weekly Product Sync - 2026-05-16]]
This is a starter note. Replace it with a real decision.
tags:
- decision
container_path: Meeting Memory > Decisions
- title: 'Sarah: update the public roadmap by Friday'
body: |
Source meeting: [[Weekly Product Sync - 2026-05-16]].
This is a follow-up note, not a live task. If you also track this in Linear, GitHub Issues, or Notion, do that there. Keep this note as the searchable memory of what was agreed.
This is a starter note. Replace it with a real follow-up.
tags:
- follow-up
container_path: Meeting Memory > Follow-ups