Databricks MCP Server

What does “Databricks” MCP Server do?

The Databricks MCP Server acts as a Model Context Protocol (MCP) server that connects AI assistants directly to Databricks environments, with a specific focus on leveraging Unity Catalog (UC) metadata. Its primary function is to enable AI agents to autonomously access, understand, and interact with Databricks data assets. The server provides tools that allow agents to explore UC metadata, comprehend data structures, and execute SQL queries. This empowers AI agents to answer data-related questions, perform database queries, and fulfill complex data requests independently, without requiring manual intervention at each step. By making detailed metadata accessible and actionable, the Databricks MCP Server enhances AI-driven development workflows and supports intelligent data exploration and management on Databricks.

List of Prompts

No specific prompt templates are mentioned in the repository or documentation.

List of Resources

No explicit list of MCP resources is provided in the repository or documentation.

List of Tools

The following tools and features are described in the documentation as being available:

• Explore Unity Catalog Metadata
  Allows AI agents to explore Databricks Unity Catalog metadata, including catalogs, schemas, tables, and columns.
• Understand Data Structures
  Enables agents to understand the structure of Databricks datasets, facilitating more accurate SQL query construction.
• Execute SQL Queries
  Provides the ability for AI agents to run SQL queries on Databricks, supporting various data requests and analysis.
• Autonomous Agent Actions
  Supports agent modes where the AI can iterate through requests and perform complex, multi-step data tasks independently.

Use Cases of this MCP Server

• Database Metadata Discovery
  AI agents can autonomously explore Databricks Unity Catalog metadata to understand data assets and relationships without manual lookup.
• Automated SQL Query Building
  Agents use metadata to automatically construct and execute SQL queries tailored to user needs or analytical tasks.
• Data Documentation Assistance
  By leveraging UC metadata, AI can assist in documenting data assets or verifying documentation completeness and accuracy.
• Intelligent Data Exploration
  Developers can use the MCP server to have AI agents answer ad hoc data questions or perform exploratory data analysis.
• Complex Task Automation
  The server’s agent mode allows AI to chain together multiple steps, such as discovering data, running queries, and returning results, all without human intervention.

How to set it up

Windsurf

No Windsurf-specific setup instructions or JSON snippets are provided.

Claude

No Claude-specific setup instructions or JSON snippets are provided.

Cursor

The repository mentions integration with Cursor:

1. Ensure you have Python and required dependencies installed.
2. Clone the repository and install requirements from requirements.txt.
3. Locate configuration files for MCP servers in Cursor.
4. Add the Databricks MCP Server to the mcpServers object:

   {
     "databricks-mcp": {
       "command": "python",
       "args": ["main.py"]
     }
   }
   

5. Save your configuration and restart Cursor if required.

Securing API Keys using Environment Variables (example):

{
  "databricks-mcp": {
    "command": "python",
    "args": ["main.py"],
    "env": {
      "DATABRICKS_TOKEN": "YOUR_API_KEY"
    }
  }
}

Cline

No Cline-specific setup instructions or JSON snippets are provided.

How to use this MCP inside flows

Using MCP in FlowHunt

To integrate MCP servers into your FlowHunt workflow, start by adding the MCP component to your flow and connecting it to your AI agent:

Click on the MCP component to open the configuration panel. In the system MCP configuration section, insert your MCP server details using this JSON format:

{
  "databricks-mcp": {
    "transport": "streamable_http",
    "url": "https://yourmcpserver.example/pathtothemcp/url"
  }
}

Once configured, the AI agent is now able to use this MCP as a tool with access to all its functions and capabilities. Remember to change “databricks-mcp” to your actual MCP server name and replace the URL with your own MCP server URL.

Overview

Based on the available documentation, the Databricks MCP Server is well-scoped for Databricks/UC integration and agentic AI workflows, but is missing explicit prompt templates, resource lists, and mentions of roots or sampling features. Its setup and tool descriptions are clear for Cursor, but less so for other platforms.

Our opinion

The MCP server is focused and useful for Databricks + AI automation, but would benefit from more explicit documentation around prompts, resources, and multi-platform setup. For those seeking Databricks/UC integration, it is a solid and practical solution.

MCP Score