#31 · Primary category: Knowledge Base & RAG

memU

agent-memory claude-skills harness loop-engineering mcp memory openclaw openclaw-skills sandbox skills

Personal memory across agents

Project last updated:08/28/26

GitHub Stars

14.4K

Forks

1.1K

Contributors

43

License

Other

Why we included this project

Coding agents lose their train of thought every time a session ends, and memU gives them somewhere to put what they learned. It runs as a lightweight sidecar alongside Claude Code, Cursor, Codex, and OpenClaw, scanning finished session logs and distilling the useful parts into a shared wiki that later tasks can consult. What stands out is the skill extraction: rather than storing raw chat history, the agent itself can patch an existing skill or create a new reusable workflow, and that knowledge gets indexed so a future session can pull it back. You can self-host it with your own storage and embedding provider or use the managed memu.so service if you want cross-device sync. For teams and solo developers running long-lived projects, it's a practical way to carry hard-won conventions across unrelated conversations without keeping notes by hand.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category