Databricks MCP integration for AI agents.
Connect AI agents to Databricks through 36 structured actions, including get cluster, create cluster, cancel run, and put secret. Review authentication, inpu…
- ACTION CONTRACTS
- 36
- AUTHENTICATION
- API key
- SOURCE STATUS
- Provider-backed
CATALOG SOURCE REVIEWED AUGUST 23, 2026 / ACTION NAMES AND SCHEMAS DERIVED FROM WORKING MACHINES PROVIDER SOURCE
CAPABILITY PROFILE
What the Databricks integration exposes
The Databricks Agent App exposes 36 provider-backed actions for data and developer tools work. Its current contract lets an authorized agent get one databricks cluster by cluster id, create a databricks cluster from a raw clusters/create payload, and cancel a databricks job run by run id. These operations are called through Working Machines as typed capabilities rather than through browser navigation or copied UI steps.
A connection uses API key. Before execution, the agent can inspect the selected action, its required fields, declared scopes, and expected output contract. Provider credentials remain inside the Working Machines runtime; the calling agent receives the capability and its structured result, not the underlying secret.
For reliable operation, start with the narrowest action that satisfies the task, resolve stable provider identifiers before changing state, and validate the returned object or status after execution. Availability still depends on the connected Databricks account, granted provider permissions, workspace policy, region, plan, and upstream API behavior.
VERIFIED ACTION SAMPLE
Real Databricks capabilities.
Showing 12 of 36 actions. Risk labels are conservative signals based on operation names, not substitutes for provider documentation or runtime policy.
get_clusterREADGet one Databricks cluster by cluster ID.
INPUTS: clusterId
create_clusterWRITECreate a Databricks cluster from a raw clusters/create payload.
INPUTS: cluster
cancel_runHIGH IMPACTCancel a Databricks job run by run ID.
INPUTS: runId
put_secretREVIEWCreate or overwrite a Databricks secret value.
INPUTS: scope / key / stringValue / bytesValue
create_jobWRITECreate a Databricks job from a raw Jobs API settings object.
INPUTS: settings
create_repoWRITECreate a Databricks workspace repo linked to a Git remote.
INPUTS: url / path / provider / branch / tag
create_secret_scopeWRITECreate a Databricks secret scope.
INPUTS: scope / scopeBackendType / backendAzureKeyvault / initialManagePrincipal
delete_jobHIGH IMPACTDelete a Databricks job by ID.
INPUTS: jobId
delete_repoHIGH IMPACTDelete a Databricks workspace repo by repo ID.
INPUTS: repoId
delete_secretHIGH IMPACTDelete one Databricks secret value.
INPUTS: scope / key
delete_secret_scopeHIGH IMPACTDelete a Databricks secret scope by scope name.
INPUTS: scope
edit_clusterWRITEEdit an existing Databricks cluster by cluster ID.
INPUTS: clusterId / cluster
INPUT CONTRACTS
Know what the action needs before it runs.
get_cluster
clusterIdREQUIRED- The Databricks cluster ID.
create_cluster
clusterREQUIRED- A raw Databricks API object.
cancel_run
runIdREQUIRED- The Databricks run ID.
put_secret
scopeREQUIRED- The Databricks secret scope name.
keyREQUIRED- The Databricks secret key name.
stringValueOPTIONAL- UTF-8 secret value.
bytesValueOPTIONAL- Bytes secret value.
PROVIDER-SPECIFIC WORKFLOWS
Jobs this Agent App can support
Inspect Get Cluster
Get one Databricks cluster by cluster ID. Use this as a bounded discovery step, retain the returned identifier, and avoid expanding the read beyond the task's stated scope.
get_clusterControl Create Cluster
Create a Databricks cluster from a raw clusters/create payload. Resolve the target first, present material changes for confirmation, and make retries idempotent where the provider supports it.
create_clusterVerify with Create Cluster
Create a Databricks cluster from a raw clusters/create payload. Compare the returned provider state with the intended outcome and preserve stable IDs or canonical links in the run record.
create_clusterSAFETY BOUNDARY
Operate Databricks with explicit limits
- Authorize Databricks with API key and grant only the provider access required by the selected actions.
- 18 actions are change-capable by name. Confirm the target identity and material parameters before allowing a write.
- Treat `cancel_run`, `delete_job`, and `delete_repo` as high-impact operations and require an explicit approval boundary.
- No provider scope string is declared on the sampled actions. Verify the connected account's actual permissions in Databricks rather than assuming unrestricted access.
- After a call, inspect the structured result and execution record before reporting that the Databricks task completed successfully.
CONNECTION MODEL
API key
Databricks personal access token used with the Authorization Bearer header.
Working Machines stores provider credentials behind the execution boundary. An agent can use an authorized connection identity, but catalog discovery alone does not reveal OAuth tokens, API keys, or provider secrets.
- READ SIGNALS
- 16
- WRITE SIGNALS
- 11
- HIGH IMPACT
- 7
- REVIEW SIGNALS
- 2
EVIDENCE AND AVAILABILITY
Provider reference
Action names, input fields, authentication types, and counts on this page are generated from the Working Machines provider catalog. Provider behavior, quotas, object semantics, account eligibility, and regional availability remain governed by Databricks.