Claw Recall
Persistent, searchable memory for AI agents. When context compaction erases what your agent was just working on, Claw Recall brings it back.
Free and open source. Under $1/month to run.
# the_problem
Every AI conversation is temporary. When context compacts, when you start a new session, when you switch agents — all that context is gone. Decisions made, code reviewed, configurations discussed — lost. Claw Recall indexes everything so any agent can search what any other agent discussed.
# search_interface
Web-based search with agent filtering, keyword highlighting, and semantic search.
# data_sources
Claw Recall isn't just conversation search. It ingests from multiple sources and makes everything searchable in one place.
Agent Conversations
OpenClaw, Claude Code, and custom agents. Active + archived sessions, all indexed incrementally.
Gmail
Inbox + sent mail across multiple accounts. Smart noise filtering skips newsletters, alerts, and automated messages.
Google Drive
Docs, Sheets, PDFs, markdown files. Re-captures when files are modified. Skips code and binary files.
Slack
DMs, group DMs, public and private channels. Captures usernames, skips bot messages.
Captured Thoughts
Agents can save insights and decisions as searchable thoughts. Any agent can find what any other agent captured.
Markdown Files
Documentation, notes, and reference files discoverable alongside conversations.
# mcp_tools
6 key tools available via MCP (Model Context Protocol). Your AI agent gets native access to search, capture, and browse — no copy-pasting needed.
search_memory — search all sources with auto keyword/semantic detection browse_recent — get full transcript of recent conversations (context recovery) capture_thought — save insights so any agent can find them later browse_activity — session summaries across agents and time ranges poll_sources — trigger Gmail/Drive/Slack ingestion on demand memory_stats — database size, message counts, embedding status # documentation
Beyond conversations and external sources, your agents can also search your project documentation — playbooks, specs, guides, and reference material. Pair Claw Recall with a documentation system and your agents have both memory and reference material in one place.
Learn about the Documentation System →# features
# quick_start
# Install
git clone https://github.com/rodbland2021/claw-recall.git
cd claw-recall
pip install -r requirements.txt
# Index your conversations
python -m claw_recall.indexer --source ~/.openclaw/agents-archive/
# Search
./recall "what did we decide about the API"
# Or start the web UI + MCP server
python -m claw_recall.web --port 8765Full setup guide in the README.