Every action below is available as an MCP tool and a verb in Cerebral OS — callable from any AI agent, Claude, Cursor, Windsurf, or your own runtime via the BYOA API. All executions are governed, audited, and dry-run safe.
Cancel Run
databricks:cancel_run
Cancel a running job. The run will transition to TERMINATING then TERMINATED state.
Write
High risk
Create Cluster
databricks:create_cluster
Create a new Databricks cluster with specified configuration.
Write
Medium risk
Create Notebook
databricks:create_notebook
Create a new notebook in the workspace with optional initial content.
Write
Medium risk
Delete Notebook
databricks:delete_notebook
Delete a notebook or folder from the workspace. This action cannot be undone.
Write
High risk
Get Cluster
databricks:get_cluster
Get detailed information about a specific cluster including configuration and status.
Read
Low risk
Get Job
databricks:get_job
Get detailed information about a specific job including its configuration and settings.
Read
Low risk
Get Notebook
databricks:get_notebook
Export and retrieve notebook content in the specified format.
Read
Low risk
Get Run
databricks:get_run
Get detailed information about a specific job run including status and results.
Read
Low risk
Install Library
databricks:install_library
Install a library on a cluster. The cluster must be running for installation to begin.
Write
Medium risk
List Clusters
databricks:list_clusters
List all clusters in the workspace with their current status and configuration.
Read
Low risk
List Jobs
databricks:list_jobs
List all jobs in the workspace with their configuration and recent run status.
Read
Low risk
List Libraries
databricks:list_libraries
List all libraries installed on a specific cluster with their status.
Read
Low risk
List Notebooks
databricks:list_notebooks
List notebooks and folders in a workspace path.
Read
Low risk
List Runs
databricks:list_runs
List job runs with optional filters for job ID, state, and time range.
Read
Low risk
Run Job
databricks:run_job
Trigger a job run with optional parameters. Returns the run ID to track progress.
Write
Medium risk
Start Cluster
databricks:start_cluster
Start a terminated cluster. The cluster will transition to PENDING then RUNNING state.
Write
Medium risk
Terminate Cluster
databricks:terminate_cluster
Terminate a running cluster. This will stop all running jobs and shut down the cluster.
Write
Medium risk
Uninstall Library
databricks:uninstall_library
Uninstall a library from a cluster. The cluster must be restarted for changes to take effect.
Write
Medium risk