Databricks App

Acerca de

This template provides a robust framework to create an MCP server that runs natively on Databricks Apps. It acts as a crucial bridge, allowing AI assistants to understand and perform actions within your Databricks environment. Developers can effortlessly define custom prompts using simple markdown files and extend AI capabilities by creating powerful Python-based tools that leverage the full Databricks SDK, all while handling authentication and protocol translation automatically.

Características Principales

• 0 GitHub stars
• Local proxy manages Databricks OAuth authentication automatically for seamless integration.
• Exposes Python functions as MCP tools via `@mcp_server.tool` decorator for Databricks SDK interaction.
• Dynamically loads AI prompts from markdown files in the `prompts/` directory.
• Comprehensive testing suite including command-line and interactive web UI tests for validation.
• FastAPI-based MCP server handles HTTP requests and Server-Sent Events (SSE).

Casos de Uso

• Empower AI assistants (e.g., Claude) to execute SQL queries and retrieve data from Databricks.
• Enable AI assistants to manage Databricks resources like listing clusters or creating jobs.
• Provide AI assistants with contextually relevant information about the Databricks workspace through custom prompts.