Loading...
An MCP server enabling LLMs to interact with Databricks for SQL queries, job management, and workspace information, secured by API token.
Boost this tool
Subscribe to listing upgrades or segmented pushes.
An MCP server enabling LLMs to interact with Databricks for SQL queries, job management, and workspace information, secured by API token.
This server provides useful Databricks interaction capabilities but relies heavily on the security of the Databricks API token. It's relatively safe for read-only operations, but the run_sql_query tool introduces a moderate risk depending on how user inputs are handled.
Performance is limited by the Databricks API and the efficiency of the SQL queries. Large datasets or complex queries may result in slow response times.
Cost is primarily determined by Databricks compute resources used by SQL queries and API calls. Monitor Databricks usage to avoid unexpected charges.
npx @modelcontextprotocol/inspector python3 main.pyDATABRICKS_HOSTDATABRICKS_TOKENDATABRICKS_HTTP_PATHrun_sql_queryExecutes a SQL query against a Databricks SQL warehouse.
Potential for SQL injection and data modification if not properly sanitized.
list_jobsLists all Databricks jobs in the workspace.
Read-only operation that retrieves a list of jobs.
get_job_statusRetrieves the status of a specific Databricks job by ID.
Read-only operation that retrieves job status information.
get_job_detailsRetrieves detailed information about a specific Databricks job.
Read-only operation that retrieves job details.
API Key
This server provides useful Databricks interaction capabilities but relies heavily on the security of the Databricks API token. It's relatively safe for read-only operations, but the `run_sql_query` tool introduces a moderate risk depending on how user inputs are handled.
Autonomy level depends on the LLM's configuration and how the tools are exposed. The `run_sql_query` tool should be carefully controlled in autonomous mode.
Production Tip
Implement robust logging and monitoring to track API usage, identify errors, and detect potential security breaches.
Store the token securely, avoid committing it to version control, and consider using Databricks secrets management.
The token requires permissions to run SQL queries, list jobs, and get job statuses. Limit the scope of the token to the minimum required permissions.
Sanitize user inputs before passing them to the `run_sql_query` tool. Use parameterized queries or escaping techniques to prevent malicious code execution.
No, the server is configured for a single Databricks workspace using the environment variables. You would need to modify the code to support multiple workspaces.
Implement logging to track API calls, query execution times, and error rates. Use Databricks monitoring tools to track resource usage.
The server is single-threaded, so it does not support concurrent requests. You may need to use a process manager to run multiple instances for handling concurrent requests.
The server will return an error message indicating that the connection to the SQL warehouse failed. Ensure the warehouse is running and accessible.