Databricks Tools

This Dify tool plugin lets workflows interact with Databricks in two demo-friendly ways:

• execute SQL queries against a Databricks SQL warehouse
• trigger a Databricks Job and inspect its latest run status

Included tools

Credentials

Configure the provider with:

• : your Databricks workspace host, for example
• : a Databricks personal access token or another bearer token accepted by the workspace
• (optional): the default warehouse ID to use for SQL queries
• (optional): the default catalog for SQL queries
• (optional): the default schema for SQL queries

Databricks APIs used

• SQL Statement Execution API:
• Jobs API 2.0: and

Local testing

The core Databricks client is covered by unit tests that do not require live Databricks credentials:

Source and support

• Source repository and issue tracker: yaxuanm/dify-plugin-databricks

Notes

• This plugin is designed as a workflow tool plugin, not a native knowledge datasource.
• For demo safety, the SQL tool blocks multi-statement queries and a set of high-risk privilege-management commands.
• The plugin follows the same high-level provider/tool separation pattern as the official Snowflake plugin, but uses Databricks REST APIs instead of a direct database connector.