google-meet-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| ZALO_USER_ID | No | Target Zalo user ID to receive direct messages. | |
| GEMINI_API_KEY | No | Google Gemini API key for multimodal chat and synthesis. | |
| TELEGRAM_CHAT_ID | No | Telegram user or group chat ID for alerts. | |
| ZALO_WEBHOOK_URL | No | Webhook endpoint for forwarding Zalo alerts. | |
| TELEGRAM_BOT_TOKEN | No | Telegram Bot token created via @BotFather. | |
| ZALO_OA_ACCESS_TOKEN | No | Official Account access token for Zalo OpenAPI v3.0. | |
| GOOGLE_MEET_TOKEN_FILE | No | Path to saved Google OAuth2 user token JSON. | ~/.google_meet_mcp/token.json |
| GOOGLE_MEET_MCP_DATA_DIR | No | Root directory for sessions, slides, audio, and exports. | ~/.google_meet_mcp |
| GOOGLE_MEET_COMPANION_HOST | No | Bind host for the companion HTTP server and dashboard. | 127.0.0.1 |
| GOOGLE_MEET_COMPANION_PORT | No | Port for the companion HTTP server and dashboard. | 8765 |
| GOOGLE_MEET_CREDENTIALS_FILE | No | Path to Google OAuth2 client credentials JSON. | ~/.google_meet_mcp/credentials.json |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| meet_check_authA | Checks Google OAuth2 authentication status, tokens, credentials, and API scopes. |
| meet_setup_authA | Initiates Google OAuth 2.0 flow in the browser to authorize Google Meet, Drive, and Calendar access. Requires credentials.json from Google Cloud Console. |
| meet_start_sessionB | Starts a new durable meeting session with append-only disk buffering. Safely handles meetings up to 3+ hours without memory exhaustion or data loss. |
| meet_ingest_chunkB | Ingests a live spoken line/chunk into an active session in real-time. Automatically flushes to disk immediately. |
| meet_get_live_statusB | Retrieves live status of a meeting: duration, word count, utterance count, speaker breakdown, and recent utterances. |
| meet_list_sessionsC | Lists all locally stored meeting sessions with metadata. |
| meet_list_recent_conferencesC | Queries Google Meet REST API v2 for recent conference records. |
| meet_fetch_google_transcriptA | Fetches full transcript entries from Google Meet API v2 for a given conference record, resolves participant names, and creates a local durable session. |
| meet_fetch_drive_transcriptB | Finds and reads Google Meet transcript documents auto-saved in Google Drive, and parses them into a durable session. |
| meet_synthesize_meetingA | Synthesizes a 3-hour+ meeting session into a comprehensive Executive Meeting Minutes report with Timeline Milestones, Decision Log, Action Items Matrix (PIC & Deadlines), and Participant Analytics. Also returns optimized AI prompts for Antigravity. |
| meet_export_minutesA | Exports the synthesized meeting minutes to a local Markdown (.md), HTML (.html, print/PDF ready), or structured JSON (.json) document in the exports directory. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 11 tools
Every tool targets a distinct action and resource: auth setup vs. auth check, live session ingestion vs. status, local session listing vs. cloud conference listing, and Meet transcript fetching vs. Drive transcript fetching. The only borderline pair is list_sessions and list_recent_conferences, but their descriptions clearly separate local storage from the Meet REST API.
All tools use a consistent meet_ prefix followed by verb_noun names in snake_case, such as check_auth, start_session, ingest_chunk, list_sessions, fetch_google_transcript, and export_minutes. There are no mixed casing styles, vague generic verbs, or unexpected naming deviations.
With 11 tools, the set is well-scoped for a Google Meet transcription and minutes-generation server. Each tool covers a meaningful step in the workflow without redundancy or unnecessary bloat.
The core workflow is well covered: auth, live ingestion, status, listing, transcript retrieval from both Meet and Drive, synthesis, and export. Minor gaps exist around session lifecycle management—there is no explicit stop/end session, delete session, or raw transcript export—but agents can work around these by using synthesize and export.