
chat-recall
Your assistant forgets. Your sessions do not. One memory across every AI coding tool you use. chat-recall indexes your Claude Code, Codex, Gemini CLI, OpenCode and Cursor sessions into a single searchable history, so an assistant can pick up where you left off instead of asking you to re-explain. - **Resume** — what was finished, what is still pending, which files were touched, what it cost - **Search your real history** — sessions, plans, tasks, CLAUDE.md files, shell history and agent notes, including what you actually typed - **See what changed** — per-file diffs and the git commits from a session, reconstructed from the transcript - **Remember decisions** — a temporal knowledge graph, so "what did we decide in March, and does it still hold" has an answer - **Catch leaked secrets** — credentials pasted into old conversations, still live, ranked for rotation The cross-tool part is the point: a single tool's own history stops at its boundary; this does not. A local CLI reads the transcripts your AI tools already write to disk, redacts secrets on your machine, and syncs. Project paths are hashed by default. You choose the scope — exclude a path or a whole tool, invert it to an allowlist, set a retention window, or delete one conversation. Self-hosting is free for one person.
First seen 2 Oct 2026. Evidence as of 5 Oct 2026.
Tools
No tool list captured yet.
Change history
No changes since the first observation. The first snapshot is the baseline.
| Source | Listing | First seen | Last seen | Versions |
|---|---|---|---|---|
| Smithery | munhq/chat-recall | 2 Oct 2026 | 5 Oct 2026 | 1 |