Redd-Archiver provides a multi-platform archiving solution with an extensive REST API for accessing and managing archives. The MCP server integration enables AI-assisted interactions and programmatic access to posts, comments, users, statistics, searches, and exports via 30+ endpoints. You can run the server using the included MCP setup, then connect Claude or other AI tools to the API to perform tasks such as querying archives, filtering by subreddit or subverse, and performing full-text searches backed by PostgreSQL. The server supports both local and production deployments and can be combined with Docker for isolated environments or Tor/HTTPS deployment as described in the project’s deployment guides.

Prerequisites:

• Python 3.7+ installed on your system
• PostgreSQL installed and configured (database, user, and permissions as required by the app)
• Git and basic command-line tools
• Optional: Docker if you prefer containerized deployment

Installation steps:

1. Clone the repository or obtain the MCP server directory for reddarchiver git clone https://github.com/online-archives/redd-archiver.git cd redd-archiver

2. Set up a Python environment and install dependencies python3 -m venv venv source venv/bin/activate pip install -r requirements.txt

3. Configure PostgreSQL database

   • Create a database (e.g., redd_archiver)
   • Create a user and grant privileges as required by the application
   • Ensure the application can connect to PostgreSQL (update config/env as needed)

4. Run the MCP server using the MCP runner (example using uv) uv --directory /path/to/mcp_server run python server.py

   Or adapt to your environment, using the provided mcp_config example for reddarchiver

5. (Optional) Run via Docker

   • Build or pull the appropriate Docker image for reddarchiver if provided by the project
   • Start the container with appropriate environment variables and port mappings

6. Verify startup

   • Access the API endpoints at the configured host/port
   • Check logs for any startup errors and ensure PostgreSQL connectivity

Tips and notes:

• The README indicates the MCP server is designed for AI integration and rich REST API access; ensure your AI tool (e.g., Claude) can reach the API URL you configure (REDDARCHIVER_API_URL).
• PostgreSQL full-text indexing is a core feature; ensure the database is properly configured for FTS to get best search performance.
• If deploying behind Tor or HTTPS, refer to the deployment guides linked in the repository (Tor Hidden Service, HTTPS Production).
• When using the uv runner, you may need to adjust the --directory path to point to your actual mcp_server directory and server.py entrypoint.
• Review environment variables carefully; placeholders like REDDARCHIVER_API_URL should be updated to your deployment URL or left as a local development URL during testing.