OpenClawBrain
Install

Add trustworthy memory to your local agents.

OpenClawBrain installs as an OpenClaw plugin. It gives agents continuity, stale-memory judgment, local proof, and the Codex continuity commands without modifying OpenClaw core.

Install

openclaw plugins install clawhub:openclawbrain@0.2.33 --force
openclaw plugins enable openclawbrain
openclaw gateway restart

Requires OpenClaw 2026.5.2 or later.

Codex continuity commands

/brain codex status
/brain codex threads
/brain codex last --latest
/brain codex messages --latest --limit 5
/brain codex bind <thread-id>
/brain codex tail --bound
/brain codex note The failure looks like the auth mock.
/brain codex act --with-notes Please fix only the focused test.
/brain codex steer --bound Stop after this test run and report blockers.
/brain codex detach
/brain codex handoff --bound

Use these from OpenClaw or Telegram when Codex UI is doing the deep work at the computer. Status gives the pulse, threads finds the right goal, last/messages copy what Codex just said, tail follows briefly, note adds passive context, act sends a real instruction, steer redirects an active turn, detach stops forwarding, and handoff gives a return-to-computer brief.

Memory Graph Maintenance runs automatically after install: passive health/proposal cycles start in the background, and only deterministic low-risk repairs can apply without review.

The public bridge is safe by default. Trusted local profiles can enable Telegram-to-Codex writes and steering with exact binding, trusted senders, write allowlists, and high-risk refusal.

Graph maintenance commands

/brain graph health
/brain graph dry-run
/brain graph proposals
/brain graph explain <proposal-id>
/brain graph apply <proposal-id>
/brain graph reject <proposal-id>
/brain graph stale
/brain graph clusters
/brain graph tombstones

Use these to inspect memory graph health, generate dry-run cleanup proposals, review duplicate clusters, flag stale high-authority memories, and apply only low-risk deterministic repairs.

Tiny proof loop

openclaw plugins inspect openclawbrain --runtime
openclaw doctor
# /plugins/openclawbrain/proof?limit=10
# /plugins/openclawbrain/search?query=pnpm&limit=10
# /plugins/openclawbrain/route-policy
# /plugins/openclawbrain/graph/health
# /plugins/openclawbrain/graph/dry-run

The README covers config, Ollama, multi-agent scope, authenticated routes, route-policy-v3 learning, and debugging.