MIND-Mem
Governed replayable memory layer for AI agents with hybrid recall and audit evidence.
Tested on Linux
Use it
Linux x86_64
install mind-mem with MCP extra via uv tool install
uv tool install "mind-mem[mcp]" --force
Where the README and reality differ
- README: pip install mind-mem provides mind-mem-mcp
What worked: mind-mem-mcp requires extra: pip install "mind-mem[mcp]" (uv tool install "mind-mem[mcp]") else error No module named fastmcp
Check that it works
mind-mem-mcp --helpThe output should contain “usage: mind-mem-mcp”.
First thing to try
mind-mem-init ~/my-workspace && ls ~/my-workspace && cat ~/my-workspace/mind-mem.jsonAdd it to your agent
claude mcp add mind-mem -- mind-mem-mcpJSON for Claude Desktop, Cursor and other clients
{
"mcpServers": {
"mind-mem": {
"command": "mind-mem-mcp",
"args": []
}
}
}Run mind-mem-init <workspace> to create a workspace with a mind-mem.json config, then connect the MCP server to your AI client.
What it does
- Give AI agents memory that persistsNot shown working
- Recall memory with audit evidenceNot tested
- Search memory with hybrid BM25 and vector searchNot shown working
- Find drift and contradictions in memoryNot shown working
How far pickfit got
Linux x86_64
- Installs
- Connects
- 0 of 4 features work
- Works offline
Good for
- recall past decisions with hybrid BM25F search
- detect drift and contradictions in memory
An alternative to Mem0, claude-mem and Zep.
How pickfit tested it
Each run starts from a clean machine and follows the steps above. How pickfit verifies
Linux x86_64
An MCP client called one of its tools
- Replayed from scratch
- Works offline
- Checked on 1 Nov 2026
- Version 5.0.3
An MCP client handshake over stdio listed 107 tools, and a call to recall returned a result.
Differences the test machine hit
- README: MIND_MEM_WORKSPACE env var implied
In the test: MCP server requires MIND_MEM_WORKSPACE env var pointing to initialized workspace to start