What is the main purpose of this Databricks tool?

It exposes Databricks Unity Catalog metadata and data to AI agents in an LLM-friendly format, enhancing automated data discovery, SQL generation, and pipeline analysis by providing deterministic data access.

What specific data operations can AI agents perform?

Agents can discover catalogs and schemas, list tables, fetch detailed table metadata (columns, constraints, lineage), and execute read-only Spark SQL queries against Databricks SQL Warehouse.

Is the tool secure and can it modify data?

It offers secure, environment-driven authentication (PAT or OAuth) and adheres to the principle of least privilege. All SQL query execution is intentionally read-only (DQL) to ensure safe, auditable data retrieval.

How does it benefit AI agents and LLMs?

It provides clearer data context, improves SQL query precision, enables faster data exploration, and offers lineage-aware reasoning by converting complex UC metadata into compact, LLM-optimized Markdown summaries.

Databricks

Name: Databricks
Author: vinay-ram1999

0•

Exposes Databricks Unity Catalog metadata and data to AI agents in a deterministic, LLM-friendly format.

This project implements an MCP server that converts complex Databricks Unity Catalog (UC) and lineage metadata into compact Markdown summaries, specifically optimized for consumption by Large Language Models (LLMs) and agents. It provides a small REST adapter with environment-driven authentication (PAT or OAuth client-credentials) and exposes a concise set of async tools for agents to discover catalogs/schemas, inspect table metadata and lineage, and execute read-only SQL queries. The goal is to enhance automated data discovery, SQL generation, and pipeline analysis by providing AI agents with clear, structured data context directly from Unity Catalog.

Key Features

01Offers tools to fetch schemas, tables, and detailed table information

02Executes read-only Spark SQL queries against Databricks SQL Warehouse

030 GitHub stars

04Designed for agent composition with small, deterministic tools

05Converts UC and lineage metadata into LLM-optimized Markdown summaries

06Provides environment-driven authentication (PAT or OAuth client-credentials)

Use Cases

01Automated data discovery and exploration by AI agents

02Improved SQL generation and query precision using UC metadata

03Lineage-aware reasoning for pipeline analysis and validation

04Programmatically finding candidate tables and columns for data retrieval