meeting-memory-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@meeting-memory-mcp Log today's standup and check for any contradictions"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
meeting-memory-mcp
A local-first MCP server that gives Claude memory across your meeting transcripts.
Most "AI meeting notes" tools summarize one meeting at a time and forget everything once the chat window closes. This server stores transcripts locally (SQLite, full-text search) and lets Claude maintain a living, human-readable markdown wiki of decisions, open commitments, and contradictions across meetings — so the next time you talk to Claude about a project or a client, it already remembers what happened last time.
Everything stays on your machine. No transcript or summary is ever sent anywhere except to the LLM you're already chatting with (Claude).
How it works
You paste a transcript into Claude and say "log this meeting."
Claude calls
add_meetingto store it, thensearch_meetingsto pull up anything related from past meetings.Claude reasons over both (in its own context — this server does no LLM calls itself) and calls
write_wiki_pageto update a markdown page for that topic/client/project with the latest decisions, status, and any contradictions it noticed.Next time, Claude calls
read_wiki_pagefirst to recall context before you even mention details.
The wiki pages are plain markdown files on disk (~/.meeting-memory-mcp/wiki/ by
default) — you can open, edit, or read them yourself anytime, no app required.
Related MCP server: kb
Tools
Tool | Purpose |
| Store a new transcript |
| Full-text search across all past transcripts |
| Fetch a full transcript by id |
| List all stored meetings |
| Create/update a durable markdown memory page |
| Read a memory page |
| List all memory pages |
Setup
Requires uv.
git clone <this repo>
cd meeting-memory-mcp
uv syncClaude Desktop / Claude Code
Add to your MCP config (claude_desktop_config.json or .mcp.json):
{
"mcpServers": {
"meeting-memory": {
"command": "uv",
"args": ["--directory", "/absolute/path/to/meeting-memory-mcp", "run", "meeting-memory-mcp"]
}
}
}Restart Claude. Then just say: "Here's a transcript from today's call, log it and let me know if it conflicts with anything we discussed before."
Where data is stored
Defaults to ~/.meeting-memory-mcp/. Override with the MEETING_MEMORY_DIR environment
variable (e.g. to keep separate memory per client or sync the folder via your own
backup tool).
Why this exists
Every meeting-notes tool on the market summarizes a single meeting well. None of them help a non-technical user notice "you committed to this three meetings ago and it never came up again," because that requires reasoning across history, not just one transcript. This server doesn't do that reasoning itself — it gives Claude the storage and retrieval it needs to do it, and a human-readable place to keep what it learns.
License
MIT
Available Tools
7 toolsadd_meetingA
Store a new meeting transcript locally so it can be searched and referenced in future sessions. Call this whenever the user shares a new meeting transcript.
Args: title: Short title for the meeting, e.g. 'Acme Corp - pricing call' transcript: The full raw transcript text meeting_date: Date of the meeting, ISO format if known (e.g. 2026-06-24)
Inputs: title (required), transcript (required), meeting_date. Returns: result
get_meetingA
Retrieve the full raw transcript text of a previously stored meeting by its id (returned by search_meetings or list_meetings).
Args: id: The meeting id
Inputs: id (required). Returns: result
list_meetingsA
List all stored meetings with id, title, and date, most recent first.
Inputs: no named parameters. Returns: result
list_wiki_pagesA
List all existing wiki pages with their slug, title, and last-updated time, most recently updated first. Use this to see what topics already have durable memory before deciding whether to create a new page or update an existing one.
Inputs: no named parameters. Returns: result
read_wiki_pageA
Read the current content of a wiki page by slug. Call this at the start of a session about a known topic to recall prior context before reading raw transcripts.
Args: slug: The page's stable identifier
Inputs: slug (required). Returns: result
search_meetingsA
Full-text search across all previously stored meeting transcripts. Use this before summarizing a new meeting to find related history, prior commitments, or context that might contradict the current discussion.
Args: query: Search terms, e.g. 'pricing' or 'renewal date' limit: Maximum number of results to return (1-50)
Inputs: query (required), limit. Returns: result
write_wiki_pageA
Create or fully overwrite a persistent local wiki page (markdown) that summarizes the durable knowledge about a topic, person, project, or recurring meeting series: key decisions, current status, open commitments, and any contradictions found across meetings. This is the long-term memory that future sessions will read back. Always start the content with a '# Title' heading.
Args: slug: A short stable identifier for this page, e.g. 'acme-corp' or 'q3-roadmap' content: Full markdown content of the page, starting with '# Title'
Inputs: slug (required), content (required). Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 7 tools
Tools clearly separate meeting operations (add_meeting, get_meeting, list_meetings, search_meetings) from wiki operations (write_wiki_page, read_wiki_page, list_wiki_pages). Within each domain, each tool serves a distinct purpose with no overlap.
All tool names follow a consistent verb_noun pattern with snake_case (add_meeting, get_meeting, list_meetings, search_meetings, write_wiki_page, read_wiki_page, list_wiki_pages). No deviations or mixing of conventions.
7 tools is a well-scoped set covering two domains (meetings and wiki). Each tool provides essential functionality without unnecessary redundancy, fitting the server's purpose of storing and retrieving meeting transcripts and wiki summaries.
The tool surface covers core CRUD operations for meetings (add, get, list, search) and wiki (write, read, list). Missing delete and update for meetings, and delete for wiki, but these are minor gaps; agents can work around by writing new wiki pages or relying on search.
Related MCP Connectors
Persistent, governed institutional memory for Claude Code — specs, decisions, learnings.
Persistent memory for AI assistants. Save once; recall from Claude, ChatGPT, or any MCP client.
Persistent AI memory shared across Claude, ChatGPT, coding agents, and compatible MCP clients.
Hosted MCP memory: save sessions/decisions once, search from Claude, Cursor, ChatGPT. EU-hosted, keyword search.
Related MCP Servers
- AlicenseBqualityBmaintenanceAn MCP server that provides searchable local storage for Claude conversation history, featuring automatic topic extraction and weekly insight summaries. It enables Claude to retrieve context from past sessions through full-text search and organized file storage.103MIT
- AlicenseNot gradedqualityDmaintenanceA local MCP server that provides a shared context and learning foundation across multiple AI tools (Claude, Copilot, Codex) for multiple projects, enabling persistent knowledge, decisions, and gap reflection through note storage.48 PyPIMIT
- AlicenseNot gradedqualityDmaintenancePersistent memory MCP server for Claude Code that captures and recalls project context across sessions, eliminating the need to re-explain architecture and decisions daily.1MIT
- AlicenseNot gradedqualityDmaintenanceMCP server providing Claude with persistent, local memory for tracking architectural decisions, dead ends, and project context across conversations.MIT