Attendee

This MCP server lets you send and manage meeting bots using Attendee, an open-source meeting bot for recording and transcription. Run Attendee locally or deploy it to the cloud, and control bots in real meetings through natural language commands.

How to use

You interact with Attendee MCP through an MCP client. Create and manage meeting bots, have them speak or send messages in meetings, retrieve transcripts and recordings, and monitor bot status. Use natural language prompts to issue actions like sending a bot to a meeting, making the bot speak, sending chat messages, or removing the bot from the meeting. You can combine this MCP with other MCPs to build richer workflows, such as fetching information online and having the bot relay the result in the meeting.

How to install

Prerequisites: you need Node.js and npm installed on your system. You may also run this in a cloud environment if you prefer.

Step 1: Clone the Attendee MCP repository and navigate into it.

Step 2: Install dependencies.

Step 3: Build the TypeScript sources.

Step 4: Link the CLI for global use.

Step 5: Start the Attendee MCP locally when you are ready to test.

Running the server locally follows a build-and-run flow. You will eventually start the Attendee backend and connect the MCP to your client.

# Clone the fork or main repo
git clone https://github.com/rexposadas/attendee-mcp.git
cd attendee-mcp

# Install dependencies
npm install

# Build the TypeScript sources
npm run build

# Link for global use
npm link

Attendee MCP configuration for your client

Configure your MCP client to load the Attendee MCP server. The following configuration runs Attendee as a local stdio process and passes the required environment values for the backend.

{
  "mcpServers": {
    "attendee": {
      "command": "node",
      "args": [
        "<path>/attendee-mcp/dist/index.js"
      ],
      "env": {
        "MEETING_BOT_API_URL": "<attendee-url-here>",
        "MEETING_BOT_API_KEY": "your-api-key-here"
      }
    }
  }
}

Notes on running locally

If you want to test end-to-end locally, start the Attendee server and then test commands in your MCP client. A common flow is to build the project, start the server, and then verify the bot can join a meeting, speak, and respond in chat.

Configuration

Environment variables you need to configure your Attendee MCP server are shown above. Ensure the Attendee backend is reachable at the URL you provide and that you supply a valid API key.

Troubleshooting

Common issues include network errors or API key problems. Make sure the Attendee server is running at the configured URL, the API key is correct, and the environment variables MEETING_BOT_API_URL and MEETING_BOT_API_KEY are set in your shell profile.

If the MCP server does not appear in your MCP client, restart the client after config changes and verify the JSON config syntax is valid. Check client logs for error messages.

If you encounter method-not-allowed errors, verify that the corresponding endpoints are implemented on the Attendee server and review server logs for details.

License

MIT License - see LICENSE file for details.

Contributing

1. Fork the project
2. Create a feature branch: git checkout -b feature-name
3. Implement and test
4. Commit changes: git commit -m 'Add feature'
5. Push to the branch: git push origin feature-name
6. Submit a pull request

Available tools

create_meeting_bot

Create a bot that joins a meeting and starts recording and transcription.

get_bot_status

Query the current status of a running meeting bot.

remove_meeting_bot

Remove a bot from its meeting and clean up resources.

make_bot_speak

Convert text to speech and have the bot speak in the meeting.

send_chat_message

Send a chat message from the bot to the meeting chat.

send_image_to_meeting

Display an image through the bot in supported meetings (Google Meet).

send_video_to_meeting

Play a video through the bot in supported meetings (Google Meet).

get_meeting_transcript

Retrieve the transcript of the meeting the bot joined.

get_chat_messages

Fetch chat messages from the meeting.

get_recording

Obtain the download URL for the meeting recording.

delete_bot_data

Permanently delete all data associated with a bot.