databricks-mcp-genie

# Databricks MCP Genie A Model Context Protocol (MCP) server with enhanced Genie AI integration that provides seamless natural language interaction between AI assistants (like Claude Desktop, Cursor) and Databricks workspaces. ## What This Does Enables AI assistants to directly interact with your Databricks workspace: - Execute SQL queries and manage warehouses - Control clusters (create, start, stop, monitor) - Run jobs and notebooks - Ask natural language questions with Genie AI - Manage Unity Catalog (catalogs, schemas, tables) - Work with DBFS, repos, and libraries ## Quick Start ### For Cursor Users **Team Installation (Recommended)**: See [Cursor Setup Guide](docs/CURSOR_SETUP.md) for one-click installation instructions. Quick install: ```bash pip install databricks-mcp-genie ``` Then configure in Cursor settings - full details in the [Cursor Setup Guide](docs/CURSOR_SETUP.md). ### Prerequisites - Python 3.10 or higher - Databricks workspace with personal access token - Cursor IDE, Claude Desktop, or any MCP-compatible client ### Installation ```bash # Install from PyPI (recommended) pip install databricks-mcp-genie # Or install from source git clone https://github.com/sidart10/databricks-mcp-genie.git cd databricks-mcp-genie pip install -e ".[dev]" ``` ### Configuration 1. **Get your Databricks credentials:** - Workspace URL: `https://your-workspace.cloud.databricks.com` - Personal Access Token: Generate from User Settings > Developer > Access Tokens 2. **Configure MCP client** (Claude Desktop example): Edit `~/.config/Claude/claude_desktop_config.json`: ```json { "mcpServers": { "databricks": { "command": "/path/to/databrics-mcp-server/.venv/bin/python", "args": ["-m", "databricks_mcp.main"], "cwd": "/path/to/databrics-mcp-server", "env": { "DATABRICKS_HOST": "https://your-workspace.cloud.databricks.com", "DATABRICKS_TOKEN": "your-personal-access-token-here" } } } } ``` 3. **…