Repository navigation
Initial public release of MCP Databricks!
This version introduces a brand new MCP server for Databricks, designed to streamline and automate common Databricks operations via the Model Context Protocol (MCP). While this is an early, non-official release, it already provides a robust set of features for developers and data teams.
✨ Highlights
- Cluster Management
- List, create, start, delete, and inspect Databricks clusters
- SQL Warehouse Management
- List and create SQL warehouses
- Execute SQL statements
- Workspace Object Management
- Import and export workspace objects (notebooks, files, etc.)
- Delete objects and create directories
- Library Management
- Install libraries (JAR, WHL, PyPI, Maven, CRAN, etc.) on running clusters
- Command Execution
- Run Python, Scala, or SQL commands on clusters
- Create execution contexts for interactive sessions
🛠️ Getting Started
See the README.md for setup instructions, prerequisites, and usage examples.