home / mcp / databricks mcp server

Databricks MCP Server

This read-only MCP Server allows you to connect to Databricks data from Claude Desktop through CData JDBC Drivers. For full CRUD support, check out the first managed MCP platform: CData Connect AI (https://www.cdata.com/ai/).

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "cdatasoftware-databricks-mcp-server-by-cdata": {
      "command": "java",
      "args": [
        "-jar",
        "/PATH/TO/CDataMCP-jar-with-dependencies.jar",
        "/PATH/TO/databricks.prp"
      ]
    }
  }
}

You have a Model Context Protocol (MCP) server that lets you query live Databricks data through the CData JDBC Driver without writing SQL. This read-only server wraps the driver so large language models can fetch up-to-date information by asking natural language questions, with data surfaced locally on your machine.

How to use

You will run the MCP server on your machine and connect an MCP client to it. Start the server with the prepared configuration file that points to your JDBC connection details. Once running, ask your AI client questions about your Databricks data, and use built-in tools to explore tables, inspect columns, and run safe queries. This setup enables live data access for natural language queries without exposing direct SQL execution in your client.

How to install

# Prerequisites
- Java (JDK 8+)
- Maven (for building the MCP server)
- A Databricks JDBC driver installed and licensed

# 1. Build the MCP server
mvn clean install
# This creates the JAR: CDataMCP-jar-with-dependencies.jar

# 2. Prepare the JDBC connection
# Install and license the CData JDBC Driver for Databricks from the official source
# Follow the license steps provided by the driver, resulting in a JDBC URL string

# 3. Create a .prp config file for your connection
# Replace placeholders with your actual values
Prefix=databricks
ServerName=CDataDatabricks
ServerVersion=1.0
DriverPath=PATH\TO\cdata.jdbc.databricks.jar
DriverClass=cdata.jdbc.databricks.DatabricksDriver
JdbcUrl=jdbc:databricks:InitiateOAuth=GETANDREFRESH;
Tables=

# Save this as databricks.prp

Additional configuration and run steps

Use the following commands to run the MCP server locally. The server starts in stdio mode, which means it is intended to be used by a local client on the same machine.

java -jar /PATH/TO/CDataMCP-jar-with-dependencies.jar /PATH/TO/databricks.prp

Troubleshooting and tips

If you do not see the MCP server in your client, ensure you have fully restarted the client after starting the server. Verify the JDBC connection string and path to the driver jar. If data cannot be retrieved, double-check the connection string and ensure the Databricks OAuth flow completes successfully.

License

MIT License

This MCP server is licensed under the MIT License. You may use, modify, and distribute it in accordance with MIT terms.

Available tools

databricks_get_tables

Retrieves a list of tables available in the Databricks data source. The output is provided in CSV format, with the first line containing column headers.

databricks_get_columns

Retrieves a list of columns for a specified table. The output is provided in CSV format, with the first line containing column headers.

databricks_run_query

Executes a SQL SELECT query against the configured Databricks data source and returns the results.