mcpbeat Sign in

M3 Memory MCP Server

by skynetcmd Your server? Claim it
local only

M3 Memory runs on your own machine — the client starts it, so there is no endpoint to ping. 2 015 installs a week from pypi. Last commit 18 Sep 2026.

Local-first memory — 100+ tools, 99.2% LongMemEval-S retrieval@10, hybrid search, GDPR, no cloud.

Installs per day peak 4 565 · avg 737 · -33% w/w
a month agotoday
2 015
Installs / week
pypi · m3-memory
25
Stars
1 open issues
18 Sep 2026
Last commit
0 releases in 90 days
Apache-2.0
License
Python

What the code does

We read the source, 22 h ago · rules 3dff92dd89df

A tool parameter reaches a dangerous call

A value the model can set ends up inside a file or shell call. That is not a flaw by itself — for a terminal server it is the job — but it is where things go wrong when it is not.

A tool parameter reaches a dangerous call enrich_pending.dry_run → bin/m3_upgrade.py:181
        return subprocess.run(cmd, timeout=timeout).returncode  # nosec B603 - argv list, no shell
Capabilities

What this server is able to do. For an MCP server this is often the job itself — a terminal server runs commands because that is what it is for. Listed so you know what you are plugging in, not as an accusation.

_LAUNCHD_PLIST = "~/Library/LaunchAgents/com.m3memory.cognitiveloop.plist"
File ships in the package but is absent from the source [пакет] m3_memory/install_os.py:1
этот файл ставится пользователю, но в репозитории его нет
    path = os.path.join(chatlog_config.SPILL_DIR, f"{day}.jsonl")
    exec(code, {"__file__": str(bridge), "__name__": "__main__"})  # nosec B102

Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.

A tool parameter here reaches a dangerous call

That is not a flaw by itself — but it is where things go wrong when it is not the job. We re-read this code on every release. Watch it and you hear from us the day another one appears.

Three servers free · no card

Connect this server

This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.

run in your terminal
claude mcp add m3-memory -- uvx m3-memory
~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "m3-memory": {
      "args": [
        "m3-memory"
      ],
      "command": "uvx"
    }
  }
}
~/.codex/config.toml
[mcp_servers.m3-memory]
command = "uvx"
args = ["m3-memory"]
.cursor/mcp.json
{
  "mcpServers": {
    "m3-memory": {
      "args": [
        "m3-memory"
      ],
      "command": "uvx"
    }
  }
}
.vscode/mcp.json
{
  "mcpServers": {
    "m3-memory": {
      "args": [
        "m3-memory"
      ],
      "command": "uvx"
    }
  }
}

Alternatives to M3 Memory

same job, measured the same way
Superlocalmemory
by varun369

Local-first AI memory with knowledge graphs and hybrid search. 17+ AI tools via MCP. Free.

1 596 installs/wk local only
Memento
by scrypster

Persistent memory for AI tools. Local-first knowledge graph with hybrid search.

local only
Sverklo
by sverklo

Local-first MCP code intelligence: 37 tools — hybrid search, blast-radius, diff review, memory.

281 installs/wk local only
Codemunch Pro
by bigjai

Code indexing MCP: 15 tools, 10 languages, hybrid search, call graphs, O(1) retrieval.

125 installs/wk local only
Tokennuke
by bigjai

Code indexing MCP: 15 tools, 10 languages, hybrid search, call graphs, O(1) retrieval.

61 installs/wk local only
Cozo Memory
by tobs-code

Local-first memory system for AI agents with hybrid search and graph reasoning

64 installs/wk local only
Tacitus MCP Server
by ionasrobert

Long-term memory for AI agents: local-first PKM with provenance and hybrid search.

33 installs/wk local only
MCP Memory Graph
by yonasvalentin

Local-first memory for Claude Code and any MCP client: hybrid search + knowledge graph, $0/token.

84 installs/wk local only

M3 Memory — questions

Answers built from our own checks of this server.

Why is there no uptime for M3 Memory?
M3 Memory runs on your own machine over stdio — there is no network address to reach, so uptime cannot be measured for it by anyone. What can be measured is adoption: the pypi package m3-memory was installed 2 015 times last week.
How do I connect M3 Memory?
Copy the ready config from this page — we generate it for Claude Code, Claude Desktop, Codex, Cursor and VS Code, each with the file path that client actually reads. It runs locally, so the command pulls m3-memory straight from pypi; nothing to host, nothing to sign up for.
How many people use M3 Memory?
The pypi package m3-memory was installed 2 015 times in the last week. Week over week that is -33%. We show installs rather than GitHub stars on purpose: a star is a bookmark, an install is someone actually running it.
Is M3 Memory open source?
Yes — it is published under the Apache-2.0 licence, written in Python, 25 stars on GitHub and 1 open issue. The source link is on this page, so you can read exactly what it does with your data before you connect it.