MuckRock MCP Server

What is MuckRock MCP Server?

A Model Context Protocol (MCP) server for interacting with MuckRock's FOIA platform. It enables users to search, file, and track FOIA requests through an AI assistant like Claude.

How to use MuckRock MCP Server?

Install with pip install -r requirements.txt, then configure in the Claude Desktop config file (claude_desktop_config.json) by pointing the command to Python and the script path. On startup the server will prompt for MuckRock credentials, or you can set the environment variables MUCKROCK_USERNAME and MUCKROCK_PASSWORD.

Key features of MuckRock MCP Server

• Search FOIA requests and government agencies
• File new FOIA requests with auto organization selection
• Track status of submitted requests
• Manage organizations for filing requests
• Check authentication status
• View user info and organization details

Use cases of MuckRock MCP Server

• Search for specific FOIA requests (e.g., EPA records)
• File new FOIA requests via natural language prompts
• View pending request statuses
• Manage which organization to file requests under
• Test filing with test agencies before actual use

FAQ from MuckRock MCP Server

What are the requirements to run this server?

Python is required. Install dependencies with pip install -r requirements.txt. The server must be configured in the Claude Desktop config.

How do I authenticate with MuckRock?

The server prompts for credentials on startup. Alternatively, set the environment variables MUCKROCK_USERNAME and MUCKROCK_PASSWORD.

Is this software production-ready?

No. It is labeled experimental and users should test with test agencies (ID: 248) first. Use at your own risk.

How are authentication tokens handled?

Tokens are automatically refreshed during the session.