mcp-databricks

This MCP server exposes tools to interact with your Databricks workspace from a Model Context Protocol client. It connects to the Databricks REST API and enables running SQL queries against a Databricks SQL warehouse, listing all jobs in the workspace, and querying status or details for specific jobs. The available MCP tools include: run_sql_query(sql: str) to execute an SQL query and return results, list_jobs() to fetch the list of jobs, get_job_status(job_id: int) to retrieve the status of a given job, and get_job_details(job_id: int) to obtain more in-depth information about a job. To use these tools, your agent or client can pass natural language prompts or structured requests that map to these tool functions, enabling seamless integration with LLMs that support MCP.

Prerequisites:

• Python 3.7+
• Git
• Network access to Databricks API

Step-by-step installation:

1. Clone the repository

git clone https://github.com/your-org/jordineil-mcp-databricks-server.git
cd jordineil-mcp-databricks-server

2. Create and activate a virtual environment On macOS/Linux:

python -m venv .venv
source .venv/bin/activate

On Windows:

python -m venv .venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. Configure environment variables Create a .env file in the project root with:

DATABRICKS_HOST=your-databricks-instance.cloud.databricks.com
DATABRICKS_TOKEN=your-personal-access-token
DATABRICKS_HTTP_PATH=/sql/1.0/warehouses/your-warehouse-id

5. Run the server

python main.py

Optional: run the inspector (if you have the MCP inspector available):

npx @modelcontextprotocol/inspector python3 main.py

Tips and common considerations:

• Keep your Databricks token secure; do not commit .env to version control.
• Ensure the SQL warehouse is running and accessible from the host running this MCP server.
• If you rotate tokens, update DATABRICKS_TOKEN and restart the server.
• The .env variables correspond to the Databricks host, token, and the HTTP path for your SQL warehouse; incorrect values will result in connection errors.
• For security, consider restricting token scopes to SQL warehouse access required by your usage.