google_meet

Join Google Meet calls and convert live captions into transcripts.

Updated May 26, 2026
One-click install
npx skills add https://github.com/ruiyangruiyi/hermes-agent --skill google-meet-ruiyangruiyi
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: google_meet
Source: https://github.com/ruiyangruiyi/hermes-agent/tree/main/plugins/google_meet
Command: npx skills add https://github.com/ruiyangruiyi/hermes-agent --skill google-meet-ruiyangruiyi

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires playwright, websockets, pactl, paplay, system_profiler, ffmpeg, blackhole-2ch.

What problem does it solve?

This Skill lets you delegate meeting note-taking to an agent by joining Google Meet calls, capturing live captions into a transcript, and optionally speaking back in realtime when you explicitly want the bot to participate.

Core Features & Use Cases

  • Meet joining + transcription: Joins a user-provided meet.google.com URL and scrapes live captions into a persistent transcript file you can use for summaries and action items.
  • Realtime in-call voice (optional): In realtime mode, the agent can speak in the meeting via OpenAI Realtime audio streamed through a virtual audio device.
  • Local or remote node execution: Runs the bot on the gateway machine by default, or on a remote approved node host when Chrome credentials/profile must live on the user’s laptop (e.g., signed-in Mac).

Use it when you ask the agent to sit in on a meeting, take notes, summarize what happened, or turn spoken discussion into followups and tasks, with the choice to stay listen-only (transcribe) or respond out loud (realtime).

Quick Start

Enable the plugin and join a Meet for transcription by running: Hermes plugins enable google_meet and then Hermes meet join <your-meet-url>.

Frequently Asked Questions about google_meet

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate Google Meet transcription and capture live captions into a text file?

To automate Google Meet transcription, this Skill joins your meet.google.com URL and scrapes live captions via Playwright, persisting the output into a usable transcript file for later summaries and action items.

Can an AI agent speak in a Google Meet call using realtime voice audio?

Yes, an AI agent can speak in Google Meet by streaming OpenAI Realtime audio through a platform-specific virtual audio device bridge, enabling duplex voice participation when realtime mode is explicitly enabled.

Do I need a virtual audio device like BlackHole to enable realtime spoken responses in meetings?

Yes, realtime spoken responses require a virtual audio device such as BlackHole-2ch to bridge the OpenAI Realtime WebSocket audio stream into the Google Meet call on your local machine.

Can I run the meeting bot on a remote node host instead of the local gateway?

You can run the meeting bot on a remote approved node host, which is necessary when Chrome credentials or your signed-in profile must live on your laptop rather than the gateway machine.

What are the limitations of using Playwright for scraping live captions in Google Meet?

Caption scraping via Playwright requires strict meet.google.com URL validation and depends on Google Meet's live caption availability, meaning unsupported URLs or disabled caption settings will prevent transcript generation.

What is the best way to turn spoken meeting discussion into followups and tasks automatically?

The best way to turn spoken discussion into tasks is joining the Meet with this bot to capture captions into a persistent transcript, which you can then use for summaries and automated followup generation.