Databricks

关于

The Databricks MCP server acts as a bridge, allowing Large Language Models (LLMs) and AI clients to directly manage and interact with Databricks environments. It implements the Model Completion Protocol to expose a wide array of Databricks REST API functionalities as accessible tools, including cluster, job, and workspace file management, as well as Unity Catalog and SQL execution. Built with asyncio for efficient operation, it provides a robust way to integrate AI-driven automation into your Databricks workflows, empowering intelligent agents to create, monitor, and execute data and machine learning tasks.

主要功能

• Implements the Model Completion Protocol (MCP) for LLM interaction
• Integrates with Databricks REST API for comprehensive management
• Exposes extensive Databricks functionality as discoverable tools (e.g., clusters, jobs, notebooks, SQL)
• Built with asyncio for asynchronous and efficient operation
• Supports integration with popular AI clients like Cursor and Claude CLI
• 0 GitHub stars

使用案例

• Automating Databricks cluster, job, and notebook management via AI agents
• Enabling LLMs to query and manipulate Databricks data, including Unity Catalog assets
• Developing AI-powered workflows for continuous integration and deployment on Databricks