Chronulus MCP server enables running Chronulus AI forecasting and prediction agents inside Claude. The repository provides a ready-to-use MCP server named chronulus-agents that runs the chronulus_mcp Python module with an API key required via CHRONULUS_API_KEY. The README guides quickstart installation for Claude Desktop (macOS and Windows), including pip releases, installing from GitHub, Docker, and uvx deployment options. The MCP server is configured in Claude's claude_desktop_config.json under the mcpServers section, where command, args, and environment variables are specified. Example configuration shows using python -m chronulus_mcp with CHRONULUS_API_KEY, or running a container named chronulus-mcp with Docker. The documentation also demonstrates how to pair Chronulus with additional third-party MCP servers such as filesystem and fetch. It notes common spawn ENOENT errors that can be resolved by using absolute paths to Python or uvx executables. This MCP server works alongside Claude Settings and supports integration with other MCP servers for expanded workflows.