Fathom MCP Server provides a Python-based interface to interact with the Fathom AI API via the MCP framework. It exposes tools to list meetings, retrieve summaries and transcripts, list teams and team members, and manage webhooks for meeting notifications. You can access these capabilities through the MCP tooling exposed by the server (e.g., list_meetings, get_summary, get_transcript, list_teams, list_team_members, create_webhook, delete_webhook). The server requires a FATHOM_API_KEY to authenticate with the Fathom API. Once running, you can call the MCP endpoints or use the provided examples in the README to issue requests and integrate with other tools or automations.

Prerequisites:

• Python 3.10+ installed
• Access to the internet to install dependencies

1. Clone the repository and navigate to the project directory: git clone https://github.com/Dot-Fun/fathom-mcp.git cd fathom-mcp

2. Install dependencies (recommended via uvx; shown here with Python in-editor usage): uv pip install -e .

   or alternatively if you are using pip directly:

   pip install -e .

3. Create and configure environment variables:

   • Create a .env file in the project root or export the key in your shell: FATHOM_API_KEY=your_api_key_here
   • Obtain your API key from the Fathom settings page.

4. Run the MCP server: python server.py

   Alternatively, if you prefer the FastMCP orchestration:

   fastmcp run server.py

5. (Optional) Run via Claude Desktop integration (example):

   • Ensure your Claude Desktop config includes the fathom MCP server with the appropriate command and environment variables as shown in the README.

6. Verify the server is running by checking logs or hitting a test endpoint as described in the API section of the README.

Tips and tips:

• The server uses FATHOM_API_KEY for authentication. Ensure this key is kept secure and not committed to version control.
• If running HTTP, you can modify server.py to set transport to http and specify host/port as needed.
• The integration example shows using uv to install dependencies and simple Python invocation to run; you can adapt the command in mcp_config to uvx if you prefer uv-based execution.
• When deploying, consider containerizing with Docker as per the repository structure if you plan to run in production. Ensure environment variables are passed into the container (e.g., FATHOM_API_KEY).
• Familiarize yourself with the available tools in the Quick Start and API sections to craft complex queries (e.g., filtering meetings, including transcripts, summaries, action items, and CRM matches).
• For rate limits and error handling, respect the API limits and implement retries with backoff as needed.