How to Use OpenClaw Google Meet

The Google Meet plugin turns OpenClaw into a meeting participant, but the docs make the safety boundary explicit. It only joins an explicit https://meet.google.com/... URL, or creates a new Meet space through the Google Meet API and joins the returned URL.

30-second answer

Use openclaw googlemeet setup first, then join a specific Meet URL with openclaw googlemeet join. The default mode is agent: realtime transcription listens, the configured OpenClaw agent answers, and regular OpenClaw TTS speaks into Meet. bidi is available for direct realtime voice fallback, and transcribe observes without talk-back.

Transport choices

The plugin supports local Chrome, Chrome on a paired node, and Twilio. Chrome can run locally or on a paired node host. The default Chrome audio backend is BlackHole 2ch. Twilio accepts a dial-in number plus optional PIN or DTMF sequence; it cannot dial a Meet URL directly.

brew install blackhole-2ch sox
openclaw googlemeet setup
openclaw googlemeet setup --transport chrome-node --mode transcribe
openclaw googlemeet setup --transport twilio

Join and create

openclaw googlemeet join https://meet.google.com/abc-defg-hij
openclaw googlemeet create --transport chrome-node --mode agent
openclaw googlemeet create --access-type OPEN --transport chrome-node --mode agent
openclaw googlemeet create --no-join

Meet creation has two paths. With Google Meet OAuth credentials, OpenClaw can use the Meet API. Without those credentials, it can use browser fallback. For API creation, the docs mention scopes such as meetings.space.created, meetings.space.readonly, and meetings.space.settings depending on what you need.

Health and artifacts

The doctor surfaces meeting health such as captions, transcript lines, last caption speaker and text, realtime readiness, audio input and output state, and speech readiness. For post-meeting operations, the docs include commands for latest meetings, calendar events, artifacts, attendance, export, and ending active conferences when the authorized account can manage the space.

Operational cautions

There is no automatic consent announcement. If you deploy this in a business, the host or operator must handle consent and meeting policy. Also watch for manual action requirements: Chrome permission dialogs, guest admission, missing BlackHole devices, offline nodes, duplicate tabs, and account sign-in state are all real failure points.

Operator checklist

Before a live meeting, run setup, verify the selected node if using chrome-node, confirm BlackHole and SoX on the Chrome host, sign in to the browser profile, and do a test speech or test listen run. For Twilio, verify the Voice Call plugin, credentials, webhook exposure, and DTMF sequence.

The OpenClaw Playbook covers meeting automation as an operator workflow: explicit URL, explicit mode, explicit consent process, and a recovery path when the agent is present but cannot hear or speak.

Rollout plan

Treat How to Use OpenClaw Google Meet as a workflow you roll out in stages, not a switch you flip once. Start with the smallest harmless proof: a status check, dry run, local-only call, private session, or read-only inspection. Confirm the documented behavior matches your installed OpenClaw version, then write the exact commands and expected output into the workspace so the next agent does not rely on memory or vibes.

For a production runbook, document credential owner, webhook or browser dependency, allowed actions, rate limits, fallback channel, and the person who can approve live external writes. Also write down what the agent may do alone, what requires approval, and what must stop immediately. That boundary is the difference between useful autonomy and a workflow that surprises the operator at the worst possible time.

Keep one rollback note beside the guide. It can be as simple as the command to disable a plugin, the channel to pause, the config key to revert, or the owner who must approve the next run. Include the proof that tells you rollback worked, and keep it visible near the production checklist for future maintainers. Agents are most useful when recovery is obvious.

After the first live run, review the transcript or logs while the details are fresh. Look for missing prerequisites, stale assumptions, broad prompts, confusing errors, and any external side effect that should have been gated. Tighten the guide, then repeat with one wider scope. The OpenClaw Playbook is built around this operating rhythm: cautious first proof, written runbook, verified automation, then gradual autonomy once the evidence is boring.