mcp-databricks-server

mcp-databricks-server

8
databases

The Databricks MCP Server facilitates SQL execution on Databricks through a Model Context Protocol (MCP) interface, ideal for complex and agent-mode operations. It supports SQL querying, schema listing, and detailed table descriptions, enhancing data interaction capabilities in the Databricks environment.

Databricks MCP Server

This project is a Model Context Protocol (MCP) server built for executing SQL queries against Databricks using the Statement Execution API. It is capable of retrieving data through SQL requests and can perform complex task iterations in Agent mode. Key features include:

  • Execute SQL queries on Databricks
  • List available schemas and tables
  • Describe table schemas

Setup

Requires Python 3.10+. Install dependencies via pip, and configure environment variables for Databricks credentials. Ensure appropriate permissions for SQL warehouse, token permissions, and data access.

Running the Server

Operate in standalone mode or configure with Cursor for integration with AI assistants.

Available Tools

  • Execute SQL queries
  • List schemas and tables
  • Describe table schemas

Handling Long-Running Queries

Handles long-running queries by polling the Databricks API, with adjustable timeout settings.

Dependencies

Utilizes httpx, python-dotenv, mcp, and asyncio libraries for its operations.