This Databricks MCP server enables AI assistants to execute SQL queries against Databricks using the Statement Execution API. Built with Python using FastMCP, it provides tools for executing SQL queries, listing schemas and tables, and describing table schemas. The implementation handles authentication through environment variables, manages long-running queries with polling, and formats query results into readable tables. It's designed to work with Cursor and other MCP clients, making it ideal for data exploration, analysis, and complex tasks when coupled with Unity Catalog Metadata.
RafaelCartenet