Skip to main content

Databricks

Databricks connector for managing clusters, jobs, job runs, SQL warehouses, SQL statement execution, and workspace objects across any Databricks deployment (AWS, Azure, GCP).

Categories: Data, Analytics, SQL, Infrastructure

What you can ask​

  • List the 10 most recent failed jobs in the prod workspace and surface the error.
  • Run the nightly_etl job and notify #data-eng when it finishes.
  • Show me clusters with idle time >2 hours so we can shut them down.

Set up​

Databricks connects with credentials from your Databricks account. Have these values ready:

  • Workspace URL
  • Personal access token
  1. Go to Organization > Connectors and find Databricks.
  2. Click Enable and choose which tools to allow.
  3. Each member enters their own values the first time they use it.

To use one account for the whole organization, turn on Share one connection with everyone. Every member's agents then act as that account, with exactly the access it has.

Tools​

Databricks offers 48 tools. Admins choose which ones agents can use. See Tool permission groups.

Read (28)​

View and search information without changing it.

ToolWhat it does
ask_genieAsk a natural language question about your data using Databricks Genie.
execute_sql_statement_readonlyRun read-only SQL on a Databricks SQL warehouse: SELECT, WITH, SHOW, DESCRIBE, EXPLAIN, VALUES, LIST.
export_notebookExport a notebook's content from the workspace.
get_clusterGet detailed information about a specific Databricks cluster including state, configuration, and resource allocation.
get_cluster_eventsGet events for a cluster to diagnose issues.
get_current_userGet the current authenticated Databricks user and their workspace home directory.
get_genie_messageGet the status and results of a Genie message.
get_jobGet detailed information about a specific Databricks job including tasks, schedule, and configuration.
get_job_runGet detailed information about a specific job run including state, timing, and task details.
get_job_run_outputGet the output of one task run: notebook results, SQL output, logs, error traces.
get_sql_statementGet the status and results of a SQL statement execution.
get_sql_warehouseGet detailed information about a specific SQL warehouse including state, size, cluster count, and active sessions.
get_table_infoGet detailed information about a table including column definitions, types, and properties.
get_workspace_object_statusGet metadata about a workspace object including type, language (for notebooks), and timestamps.
list_catalogsList all catalogs in Unity Catalog.
list_cluster_optionsList the Spark runtime versions and node types this workspace accepts for create_cluster.
list_clustersList all clusters in the Databricks workspace.
list_genie_spacesList available Genie spaces (rooms).
list_job_runsList job runs in the Databricks workspace.
list_jobsList jobs in the Databricks workspace with optional name filter and pagination.
list_schemasList schemas within a Unity Catalog catalog.
list_serving_endpointsList all model serving endpoints in the workspace.
list_sql_warehousesList all SQL warehouses in the Databricks workspace.
list_tablesList tables within a Unity Catalog schema.
list_workspaceList objects in a Databricks workspace directory.
query_serving_endpointQuery a model serving endpoint: chat completions (messages), embeddings or classic models (inputs).
query_vector_indexQuery a vector search index using text or a vector.
validate_credentialValidate Databricks credentials.

Write (7)​

Create or send data.

ToolWhat it does
create_clusterCreate a new Databricks cluster.
create_jobCreate a new Databricks job with one or more tasks.
create_sql_warehouseCreate a new Databricks SQL warehouse.
create_vector_search_endpointCreate a vector search endpoint to host vector search indexes.
create_vector_search_indexCreate a vector search index on an endpoint.
run_job_nowTrigger an immediate run of a Databricks job.
run_notebookSubmit a one-time notebook run.

Delete or overwrite (13)​

Remove or replace existing data.

ToolWhat it does
cancel_job_runCancel an active job run.
cancel_sql_statementCancel a running SQL statement execution.
delete_jobPermanently delete a Databricks job.
delete_vector_search_endpointDelete a vector search endpoint.
delete_vector_search_indexDelete a vector search index.
delete_workspace_objectDelete a workspace object (notebook, file, or directory).
execute_sql_statementExecute a SQL statement that changes data or schema (INSERT, UPDATE, DELETE, MERGE, CREATE, DROP) on a SQL warehouse.
import_notebookCreate or overwrite a notebook in the workspace.
repair_runRe-run failed tasks in a completed job run without re-running succeeded ones.
start_clusterStart a terminated Databricks cluster.
start_sql_warehouseStart a stopped SQL warehouse.
stop_sql_warehouseStop a running SQL warehouse.
terminate_clusterTerminate a running Databricks cluster.
⌘J