google_meet

To transcribe live captions from a Google Meet call, you join the meeting URL and scrape the caption stream into a transcript. This Skill automates the caption capture process and stores the resulting text for follow-up actions.

Yes, you can deploy the transcription bot on a remote node. It supports remote-node execution and uses a lightweight WebSocket RPC protocol to control the transcription process and report status back to your local gateway.

Yes, the agent requires Playwright to automate the browser and join the meeting, and it uses a WebSocket RPC protocol for remote control and realtime status reporting. The Skill enforces platform checks for these prerequisite dependencies before execution.

Realtime mode is a feature that allows the agent to transcribe live captions and optionally speak directly into the meeting. It enables interactive participation rather than passive listening, covering both local and remote node-based deployments.

After transcribing a Google Meet call, you generate action items by processing the resulting transcript with Hermes tools. The Skill provides post-meeting follow-ups, summaries, and action items directly from the captured caption data.