--- title: "Databricks MCP server for AI agents - actionlake" description: "Give Claude 1,371 Databricks actions through one MCP server: catalog, ML, settings, IAM. Your agent never holds the credential. 100,000 calls a month free." canonical: https://actionlake.com/connectors/databricks updated: 2026-10-22 --- Connectors · Observability & analytics # Databricks MCP server for AI agents. actionlake is an MCP gateway: one remote MCP server that connects Claude and other MCP clients to Databricks and the rest of your systems. It gives your agent 1,371 Databricks actions (570 reads, 570 writes and 231 that wait for a confirmation) and holds the Databricks credential itself, so the agent never does. It paces Databricks's limits and records every call. [Connect Databricks free](https://actionlake.com/console?start=sandbox) [See pricing](https://actionlake.com/pricing) 100,000 free calls a month. No card, no email, no password: press Start a free sandbox on the next screen and keep the key it shows you. `claude mcp add --transport http actionlake https://connect.actionlake.com/mcp` the first call ``` do { action: "databricks.users.list", intent: "check which Databricks account is connected" } # List (users) · read · 1 upstream call ``` - 1,371 actions in one signed file - 570 read look, never change - 570 write create and update - 231 wait for a yes destructive, confirmed first Set up ## Connect Databricks in three steps. 1. 01 ### Add actionlake to Claude. Run the command once in Claude Code; its consent page can create a free sandbox: 100,000 calls a month, no card. In the Claude app, add a custom connector with the URL below and paste your key on the consent page. Creating a new sandbox from the Claude app is not yet tested by us. `https://connect.actionlake.com/mcp` 2. 02 ### Connect Databricks. OAuth 2.0, client credentials. Paste the app's client id and secret; actionlake fetches and renews the token, and nobody has to sign in. actionlake stores what you paste write-only, sealed under your tenant's key. - `client_id` Service principal application (client) ID - `client_secret` The service principal's OAuth secret (shown once when generated) - `account_id` Your Databricks account id - `tenant_host` Where your Databricks lives: your tenant host ``` account { op: "connections.requirements", service: "databricks" } # free: exactly what to paste, and what is found for you ``` 3. 03 ### Make the first call. Check before a write. Ask in plain words; Claude picks the action. Before anything changes, `check` validates the call and prices it, for free. ``` do { action: "databricks.users.list" } # list (users) · read · 1 upstream call check { action: "databricks.users.create", input: { … } } # free: inputs, permissions, cost and confirmations, nothing sent ``` Databricks at a glance **Sign-in** OAuth 2.0, client credentials. Paste the app's client id and secret; actionlake fetches and renews the token, and nobody has to sign in. **You provide** `client_id` `client_secret` **Setup fields** `account_id` `tenant_host` **Limits we pace to** 25 calls a second, per connection **API covered** Databricks REST API 2.0/2.1 as generated into databricks-sdk-py 0.156.0, at your Databricks instance's API base URL, which you enter when connecting **Built from** Databricks's documentation, read 22 October 2026[files.pythonhosted.org](https://files.pythonhosted.org/packages/d2/c1/631160994379b7baa4605c2a4ffa6bca374b3ef73213e43b2dfcbd6c2bb9/databricks_sdk-0.156.0-py3-none-any.whl) [docs.databricks.com](https://docs.databricks.com/api/workspace/introduction) **Service file** Version 1.0.0, signed (Ed25519) **Cached reads** 539 of 570 reads, for up to 5 minutes; a write clears what it affects **Shared entities** 3, so ask can line Databricks up with your other systems **Data** Processed and stored in Sydney, Australia. No US region yet Capabilities ## What an AI agent can do in Databricks. 1,371 actions in 41 areas. Every one carries its effect, so a key can allow the reads and hold the rest. - **570** read look, never change - **570** write create and update - **231** destructive wait for a confirmation 1. ### Catalog 171 68 read · 71 write · 32 destructive Create (catalogs) · List (catalogs) · Create (schemas) · List (schemas) 2. ### ML 119 40 read · 57 write · 22 destructive Create experiment (experiments) · Create issue (experiments) · Create logged model (experiments) · Create run (experiments) 3. ### Settings 108 42 read · 45 write · 21 destructive Create (account ip access lists) · List (account ip access lists) · Create (ip access lists) · List (ip access lists) 4. ### IAM 97 38 read · 46 write · 13 destructive List (users) · List (groups) · Me (current user) · Create (users) 5. ### Iamv2 80 40 read · 27 write · 13 destructive Create direct group member (account iam v2) · Create group (account iam v2) · Create identity visibility filter (account iam v2) · Create service principal (account iam v2) 6. ### SQL 69 26 read · 30 write · 13 destructive Create (queries) · List (queries) · List visualizations (queries) · List (query history) 7. ### Compute 64 27 read · 26 write · 11 destructive Create (clusters) · Events (clusters) · List (clusters) · List node types (clusters) 8. ### Postgres 52 21 read · 21 write · 10 destructive Create branch (postgres) · Create catalog (postgres) · Create cdf config (postgres) · Create data api (postgres) **33 more areas · 611 actions** - Marketplace **50** - Dashboards **49** - Workspace **35** - Provisioning **30** - Apps **29** - Agentkit **27** - Oauth2 **27** - Sharing **27** - Database **24** - Vectorsearch **24** - Jobs **23** - Cleanrooms **21** - Knowledgeassistants **20** - Serving **20** - Supervisoragents **19** - Billing **16** - Pipelines **16** - Bundledeployments **15** - Aisearch **14** - Bundle **14** - Files **14** - Sandbox **14** - Dataquality **11** - Disasterrecovery **10** - Networking **10** - Tags **10** - Environments **9** - Settingsv2 **9** - Agentbricks **6** - Domains **5** - Qualitymonitorv2 **5** - Aifunctions **4** - Dataclassification **4** Top actions ## The Databricks API, shaped for agents. The ids your agent types, what each does, and the one name it shares with other systems. 15 of Databricks's 15 action names are shared with at least one other connector, so an agent learns them once. | Action | What it does | Effect | Same name in | | --- | --- | --- | --- | | `databricks.users.list` List (users) | Gets details for all the users associated with a Databricks workspace. | read | `identity.user.list` 474 connectors | | `databricks.groups.list` List (groups) | Gets all details of the groups associated with the Databricks workspace. | read | `identity.group.list` 262 connectors | | `databricks.current_user.me` Me (current user) | Get details about the current method caller's identity. | read | `identity.current_user.get` 307 connectors | | `databricks.users.create` Create (users) | Creates a new user in the Databricks workspace. | write | `identity.user.create` 262 connectors | | `databricks.groups.create` Create (groups) | Creates a group in the Databricks workspace with a unique name, using the supplied group details. | write | `identity.group.create` 205 connectors | | `databricks.users.delete` Delete (users) | Deletes a user. | destructive | `identity.user.delete` 253 connectors | | `databricks.service_principals.list` List (service principals) | Gets the set of service principals associated with a Databricks workspace. | read | `identity.service_account.list` 35 connectors | | `databricks.service_principals.create` Create (service principals) | Creates a new service principal in the Databricks workspace. | write | `identity.service_account.create` 32 connectors | | `databricks.apps.list` List (apps) | Lists all apps in the workspace. | read | this connector | | `databricks.jobs.list` List (jobs) | Retrieves a list of jobs. | read | this connector | | `databricks.apps.create` Create (apps) | Creates a new app. | write | this connector | | `databricks.jobs.create` Create (jobs) | Create a new job. | write | this connector | Try it ## Prompts to paste into Claude. Each one names the Databricks actions it will call, so you know what it reads and what it changes before you send it. 1. Read read Who has access in Databricks? List every user and their role. calls `databricks.users.list` 2. Write write Create (users) in Databricks. Show me exactly what you will send before you save it. calls `databricks.users.create` Governance for Databricks ## Destructive actions wait for a yes. 231 Databricks actions are destructive. The first 10: - Delete (users) `databricks.users.delete` - Delete (groups) `databricks.groups.delete` - Delete (service principals) `databricks.service_principals.delete` - Delete (apps) `databricks.apps.delete` - Delete (jobs) `databricks.jobs.delete` - Delete (repos) `databricks.repos.delete` - Delete (tables) `databricks.tables.delete` - Unpin (clusters) `databricks.clusters.unpin` - Trash (lakeview) `databricks.lakeview.trash` - Delete (queries) `databricks.queries.delete` A destructive action runs only with `confirm: true`. Under the `human` policy the agent cannot confirm for itself: the run stops until a named person approves, in the console or from a signed link in an email. The cap on a key is refused before the call is sent, never after. do databricks.users.delete confirm_policy: **human** Delete (users)? **intent:** the user asked me to delete (users) run suspended · waiting for a named person · key _databricks-agent_ cannot confirm itself In the console, or from an emailed signed link. a key for this connector ``` account { op: "keys.create", name: "databricks-agent", allowed_actions: ["databricks.*"], cap_calls: 2000, confirm_policy: "human_for_writes", approval_rules: [ { match: "databricks.users.delete", approver: "lead@example.com" }, { match: "databricks.groups.delete", approver: "lead@example.com" }, { match: "databricks.service_principals.delete", approver: "lead@example.com" } ] } ``` Illustrative. Scoped to Databricks only, capped at 2,000 upstream calls, every write confirmed, and the destructive ones approved by a named person. Cost ## What Databricks costs on actionlake. Every Databricks action is one upstream call, the only thing actionlake bills. Searching, `check`, `cost`, `account` and any call Databricks refused with 401, 403, 429 or a 5xx are free. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000. - 1 upstream call per action all 1,371 Databricks actions - 1,371 actions page each page of a long list is one call - 539 reads can be cached a cache hit costs nothing - 100,000 calls a month, free paid plans from US$49 a month for 250,000 [See pricing](https://actionlake.com/pricing) Action index ## Every Databricks action. 1,371 actions by area: the id your agent calls, what it does, and its effect. **Catalog 171** - `databricks.catalogs.create` Create (catalogs) write - `databricks.catalogs.list` List (catalogs) read - `databricks.schemas.create` Create (schemas) write - `databricks.schemas.list` List (schemas) read - `databricks.account_metastore_assignments.create` Create (account metastore assignments) write - `databricks.account_metastore_assignments.delete` Delete (account metastore assignments) destructive - and 165 more in this area **ML 119** - `databricks.experiments.create_experiment` Create experiment (experiments) write - `databricks.experiments.create_issue` Create issue (experiments) write - `databricks.experiments.create_logged_model` Create logged model (experiments) write - `databricks.experiments.create_run` Create run (experiments) write - `databricks.experiments.delete_experiment` Delete experiment (experiments) destructive - `databricks.experiments.delete_logged_model` Delete logged model (experiments) destructive - and 113 more in this area **Settings 108** - `databricks.account_ip_access_lists.create` Create (account ip access lists) write - `databricks.account_ip_access_lists.list` List (account ip access lists) read - `databricks.ip_access_lists.create` Create (ip access lists) write - `databricks.ip_access_lists.list` List (ip access lists) read - `databricks.account_ip_access_lists.delete` Delete (account ip access lists) destructive - `databricks.account_ip_access_lists.get` Get (account ip access lists) read - and 102 more in this area **IAM 97** - `databricks.current_user.me` Me (current user) read - `databricks.groups.list` List (groups) read - `databricks.users.create` Create (users) write - `databricks.users.list` List (users) read - `databricks.access_control.check_policy` Check policy (access control) read - `databricks.account_access_control.get_assignable_roles_for_resource` Get assignable roles for resource (account access control) read - and 91 more in this area **Iamv2 80** - `databricks.account_iam_v2.create_direct_group_member` Create direct group member (account iam v2) write - `databricks.account_iam_v2.create_group` Create group (account iam v2) write - `databricks.account_iam_v2.create_identity_visibility_filter` Create identity visibility filter (account iam v2) write - `databricks.account_iam_v2.create_service_principal` Create service principal (account iam v2) write - `databricks.account_iam_v2.create_user` Create user (account iam v2) write - `databricks.account_iam_v2.create_workspace_assignment` Create workspace assignment (account iam v2) write - and 74 more in this area **SQL 69** - `databricks.queries.create` Create (queries) write - `databricks.queries.list` List (queries) read - `databricks.queries.list_visualizations` List visualizations (queries) read - `databricks.query_history.list` List (query history) read - `databricks.alerts.create` Create (alerts) write - `databricks.alerts.delete` Delete (alerts) destructive - and 63 more in this area **Compute 64** - `databricks.clusters.create` Create (clusters) write - `databricks.clusters.events` Events (clusters) read - `databricks.clusters.list` List (clusters) read - `databricks.clusters.list_node_types` List node types (clusters) read - `databricks.cluster_policies.create` Create (cluster policies) write - `databricks.cluster_policies.delete` Delete (cluster policies) destructive - and 58 more in this area **Postgres 52** - `databricks.postgres.create_branch` Create branch (postgres) write - `databricks.postgres.create_catalog` Create catalog (postgres) write - `databricks.postgres.create_cdf_config` Create cdf config (postgres) write - `databricks.postgres.create_data_api` Create data api (postgres) write - `databricks.postgres.create_database` Create database (postgres) write - `databricks.postgres.create_endpoint` Create endpoint (postgres) write - and 46 more in this area **Marketplace 50** - `databricks.consumer_fulfillments.get` Get (consumer fulfillments) read - `databricks.consumer_fulfillments.list` List (consumer fulfillments) read - `databricks.consumer_installations.create` Create (consumer installations) write - `databricks.consumer_installations.list` List (consumer installations) read - `databricks.consumer_installations.delete` Delete (consumer installations) destructive - `databricks.consumer_installations.list_listing_installations` List listing installations (consumer installations) read - and 44 more in this area **Dashboards 49** - `databricks.genie.create_message` Create message (genie) write - `databricks.genie.create_message_comment` Create message comment (genie) write - `databricks.genie.create_space` Create space (genie) write - `databricks.genie.execute_message_attachment_query` Execute message attachment query (genie) write - `databricks.genie.delete_conversation` Delete conversation (genie) destructive - `databricks.genie.delete_conversation_message` Delete conversation message (genie) destructive - and 43 more in this area **Workspace 35** - `databricks.repos.create` Create (repos) write - `databricks.repos.list` List (repos) read - `databricks.secrets.create_scope` Create scope (secrets) write - `databricks.secrets.list_acls` List acls (secrets) read - `databricks.git_credentials.create` Create (git credentials) write - `databricks.git_credentials.delete` Delete (git credentials) destructive - and 29 more in this area **Provisioning 30** - `databricks.account_credentials.create` Create (account credentials) write - `databricks.account_credentials.list` List (account credentials) read - `databricks.encryption_keys.create` Create (encryption keys) write - `databricks.encryption_keys.list` List (encryption keys) read - `databricks.account_credentials.delete` Delete (account credentials) destructive - `databricks.account_credentials.get` Get (account credentials) read - and 24 more in this area **Apps 29** - `databricks.apps.create` Create (apps) write - `databricks.apps.create_space` Create space (apps) write - `databricks.apps.create_update` Create update (apps) write - `databricks.apps.deploy` Deploy (apps) write - `databricks.apps.delete` Delete (apps) destructive - `databricks.apps.delete_app_thumbnail` Delete app thumbnail (apps) destructive - and 23 more in this area **Agentkit 27** - `databricks.agent_kit.append_session_items` Append session items (agent kit) write - `databricks.agent_kit.create_memory` Create memory (agent kit) write - `databricks.agent_kit.create_memory_store` Create memory store (agent kit) write - `databricks.agent_kit.create_session` Create session (agent kit) write - `databricks.agent_kit.clear_session_items` Clear session items (agent kit) write - `databricks.agent_kit.create_session_store` Create session store (agent kit) write - and 21 more in this area **Oauth2 27** - `databricks.account_federation_policy.create` Create (account federation policy) write - `databricks.account_federation_policy.list` List (account federation policy) read - `databricks.custom_app_integration.create` Create (custom app integration) write - `databricks.custom_app_integration.list` List (custom app integration) read - `databricks.account_federation_policy.delete` Delete (account federation policy) destructive - `databricks.account_federation_policy.get` Get (account federation policy) read - and 21 more in this area **Sharing 27** - `databricks.providers.create` Create (providers) write - `databricks.providers.list` List (providers) read - `databricks.providers.list_provider_share_assets` List provider share assets (providers) read - `databricks.providers.list_shares` List shares (providers) read - `databricks.providers.delete` Delete (providers) destructive - `databricks.providers.get` Get (providers) read - and 21 more in this area **Database 24** - `databricks.database.create_database_catalog` Create database catalog (database) write - `databricks.database.create_database_instance` Create database instance (database) write - `databricks.database.create_database_instance_role` Create database instance role (database) write - `databricks.database.create_database_table` Create database table (database) write - `databricks.database.create_synced_database_table` Create synced database table (database) write - `databricks.database.delete_database_catalog` Delete database catalog (database) destructive - and 18 more in this area **Vectorsearch 24** - `databricks.vector_search_endpoints.create_endpoint` Create endpoint (vector search endpoints) write - `databricks.vector_search_endpoints.list_endpoints` List endpoints (vector search endpoints) read - `databricks.vector_search_indexes.create_index` Create index (vector search indexes) write - `databricks.vector_search_indexes.list_indexes` List indexes (vector search indexes) read - `databricks.vector_search_endpoints.delete_endpoint` Delete endpoint (vector search endpoints) destructive - `databricks.vector_search_endpoints.get_endpoint` Get endpoint (vector search endpoints) read - and 18 more in this area **Jobs 23** - `databricks.jobs.create` Create (jobs) write - `databricks.jobs.list` List (jobs) read - `databricks.jobs.list_runs` List runs (jobs) read - `databricks.jobs.repair_run` Repair run (jobs) write - `databricks.jobs.cancel_all_runs` Cancel all runs (jobs) destructive - `databricks.jobs.cancel_run` Cancel run (jobs) destructive - and 17 more in this area **Cleanrooms 21** - `databricks.clean_room_asset_revisions.list` List (clean room asset revisions) read - `databricks.clean_room_assets.create` Create (clean room assets) write - `databricks.clean_room_assets.create_clean_room_asset_review` Create clean room asset review (clean room assets) write - `databricks.clean_room_assets.list` List (clean room assets) read - `databricks.clean_room_asset_revisions.get` Get (clean room asset revisions) read - `databricks.clean_room_assets.delete` Delete (clean room assets) destructive - and 15 more in this area **Knowledgeassistants 20** - `databricks.knowledge_assistants.create_example` Create example (knowledge assistants) write - `databricks.knowledge_assistants.create_knowledge_assistant` Create knowledge assistant (knowledge assistants) write - `databricks.knowledge_assistants.create_knowledge_source` Create knowledge source (knowledge assistants) write - `databricks.knowledge_assistants.list_examples` List examples (knowledge assistants) read - `databricks.knowledge_assistants.delete_example` Delete example (knowledge assistants) destructive - `databricks.knowledge_assistants.delete_knowledge_assistant` Delete knowledge assistant (knowledge assistants) destructive - and 14 more in this area **Serving 20** - `databricks.serving_endpoints.create` Create (serving endpoints) write - `databricks.serving_endpoints.create_provisioned_throughput_endpoint` Create provisioned throughput endpoint (serving endpoints) write - `databricks.serving_endpoints.list` List (serving endpoints) read - `databricks.serving_endpoints.query` Query (serving endpoints) write - `databricks.serving_endpoints.build_logs` Build logs (serving endpoints) read - `databricks.serving_endpoints.delete` Delete (serving endpoints) destructive - and 14 more in this area **Supervisoragents 19** - `databricks.supervisor_agents.create_example` Create example (supervisor agents) write - `databricks.supervisor_agents.create_supervisor_agent` Create supervisor agent (supervisor agents) write - `databricks.supervisor_agents.create_tool` Create tool (supervisor agents) write - `databricks.supervisor_agents.list_examples` List examples (supervisor agents) read - `databricks.supervisor_agents.delete_example` Delete example (supervisor agents) destructive - `databricks.supervisor_agents.delete_supervisor_agent` Delete supervisor agent (supervisor agents) destructive - and 13 more in this area **Billing 16** - `databricks.budget_policy.create` Create (budget policy) write - `databricks.budget_policy.list` List (budget policy) read - `databricks.budgets.create` Create (budgets) write - `databricks.budgets.list` List (budgets) read - `databricks.budget_policy.delete` Delete (budget policy) destructive - `databricks.budget_policy.get` Get (budget policy) read - and 10 more in this area **Pipelines 16** - `databricks.pipelines.clone` Clone (pipelines) write - `databricks.pipelines.create` Create (pipelines) write - `databricks.pipelines.list_pipeline_events` List pipeline events (pipelines) read - `databricks.pipelines.list_pipelines` List pipelines (pipelines) read - `databricks.pipelines.apply_environment` Apply environment (pipelines) write - `databricks.pipelines.delete` Delete (pipelines) destructive - and 10 more in this area **Bundledeployments 15** - `databricks.bundle_deployments.create_deployment` Create deployment (bundle deployments) write - `databricks.bundle_deployments.create_version` Create version (bundle deployments) write - `databricks.bundle_deployments.list_deployments` List deployments (bundle deployments) read - `databricks.bundle_deployments.list_operations` List operations (bundle deployments) read - `databricks.bundle_deployments.complete_version` Complete version (bundle deployments) write - `databricks.bundle_deployments.delete_deployment` Delete deployment (bundle deployments) destructive - and 9 more in this area **Aisearch 14** - `databricks.ai_search.create_endpoint` Create endpoint (ai search) write - `databricks.ai_search.create_index` Create index (ai search) write - `databricks.ai_search.list_endpoints` List endpoints (ai search) read - `databricks.ai_search.list_indexes` List indexes (ai search) read - `databricks.ai_search.delete_endpoint` Delete endpoint (ai search) destructive - `databricks.ai_search.delete_index` Delete index (ai search) destructive - and 8 more in this area **Bundle 14** - `databricks.bundle.create_deployment` Create deployment (bundle) write - `databricks.bundle.create_operation` Create operation (bundle) write - `databricks.bundle.create_version` Create version (bundle) write - `databricks.bundle.list_deployments` List deployments (bundle) read - `databricks.bundle.complete_version` Complete version (bundle) write - `databricks.bundle.delete_deployment` Delete deployment (bundle) destructive - and 8 more in this area **Files 14** - `databricks.dbfs.add_block` Add block (dbfs) write - `databricks.dbfs.create` Create (dbfs) write - `databricks.dbfs.list` List (dbfs) read - `databricks.dbfs.put` Put (dbfs) write - `databricks.dbfs.close` Close (dbfs) write - `databricks.dbfs.delete` Delete (dbfs) destructive - and 8 more in this area **Sandbox 14** - `databricks.sandbox.create_directory` Create directory (sandbox) write - `databricks.sandbox.create_sandbox` Create sandbox (sandbox) write - `databricks.sandbox.execute_command_sync` Execute command sync (sandbox) write - `databricks.sandbox.list_commands` List commands (sandbox) read - `databricks.sandbox.delete_path` Delete path (sandbox) destructive - `databricks.sandbox.delete_sandbox` Delete sandbox (sandbox) destructive - and 8 more in this area **Dataquality 11** - `databricks.data_quality.create_monitor` Create monitor (data quality) write - `databricks.data_quality.create_refresh` Create refresh (data quality) write - `databricks.data_quality.list_monitor` List monitor (data quality) read - `databricks.data_quality.list_refresh` List refresh (data quality) read - `databricks.data_quality.cancel_refresh` Cancel refresh (data quality) destructive - `databricks.data_quality.delete_monitor` Delete monitor (data quality) destructive - and 5 more in this area **Disasterrecovery 10** - `databricks.disaster_recovery.create_failover_group` Create failover group (disaster recovery) write - `databricks.disaster_recovery.create_stable_url` Create stable url (disaster recovery) write - `databricks.disaster_recovery.list_failover_groups` List failover groups (disaster recovery) read - `databricks.disaster_recovery.list_stable_urls` List stable urls (disaster recovery) read - `databricks.disaster_recovery.delete_failover_group` Delete failover group (disaster recovery) destructive - `databricks.disaster_recovery.delete_stable_url` Delete stable url (disaster recovery) destructive - and 4 more in this area **Networking 10** - `databricks.endpoints.create_endpoint` Create endpoint (endpoints) write - `databricks.endpoints.list_endpoints` List endpoints (endpoints) read - `databricks.private_network_gateways.create_private_network_gateway` Create private network gateway (private network gateways) write - `databricks.private_network_gateways.list_private_network_gateways` List private network gateways (private network gateways) read - `databricks.endpoints.delete_endpoint` Delete endpoint (endpoints) destructive - `databricks.endpoints.get_endpoint` Get endpoint (endpoints) read - and 4 more in this area **Tags 10** - `databricks.tag_policies.create_tag_policy` Create tag policy (tag policies) write - `databricks.tag_policies.list_tag_policies` List tag policies (tag policies) read - `databricks.workspace_entity_tag_assignments.create_tag_assignment` Create tag assignment (workspace entity tag assignments) write - `databricks.workspace_entity_tag_assignments.list_tag_assignments` List tag assignments (workspace entity tag assignments) read - `databricks.tag_policies.delete_tag_policy` Delete tag policy (tag policies) destructive - `databricks.tag_policies.get_tag_policy` Get tag policy (tag policies) read - and 4 more in this area **Environments 9** - `databricks.environments.create_workspace_base_environment` Create workspace base environment (environments) write - `databricks.environments.get_default_workspace_base_environment` Get default workspace base environment (environments) read - `databricks.environments.get_operation` Get operation (environments) read - `databricks.environments.list_workspace_base_environments` List workspace base environments (environments) read - `databricks.environments.delete_workspace_base_environment` Delete workspace base environment (environments) destructive - `databricks.environments.get_workspace_base_environment` Get workspace base environment (environments) read - and 3 more in this area **Settingsv2 9** - `databricks.account_settings_v2.get_public_account_setting` Get public account setting (account settings v2) read - `databricks.account_settings_v2.list_account_settings_metadata` List account settings metadata (account settings v2) read - `databricks.account_settings_v2.list_account_user_preferences_metadata` List account user preferences metadata (account settings v2) read - `databricks.workspace_settings_v2.list_workspace_settings_metadata` List workspace settings metadata (workspace settings v2) read - `databricks.account_settings_v2.get_public_account_user_preference` Get public account user preference (account settings v2) read - `databricks.account_settings_v2.patch_public_account_setting` Patch public account setting (account settings v2) write - and 3 more in this area **Agentbricks 6** - `databricks.agent_bricks.create_custom_llm` Create custom llm (agent bricks) write - `databricks.agent_bricks.get_custom_llm` Get custom llm (agent bricks) read - `databricks.agent_bricks.start_optimize` Start optimize (agent bricks) write - `databricks.agent_bricks.update_custom_llm` Update custom llm (agent bricks) write - `databricks.agent_bricks.cancel_optimize` Cancel optimize (agent bricks) destructive - `databricks.agent_bricks.delete_custom_llm` Delete custom llm (agent bricks) destructive **Domains 5** - `databricks.domains.create_domain` Create domain (domains) write - `databricks.domains.get_domain` Get domain (domains) read - `databricks.domains.list_domains` List domains (domains) read - `databricks.domains.update_domain` Update domain (domains) write - `databricks.domains.delete_domain` Delete domain (domains) destructive **Qualitymonitorv2 5** - `databricks.quality_monitor_v2.create_quality_monitor` Create quality monitor (quality monitor v2) write - `databricks.quality_monitor_v2.get_quality_monitor` Get quality monitor (quality monitor v2) read - `databricks.quality_monitor_v2.list_quality_monitor` List quality monitor (quality monitor v2) read - `databricks.quality_monitor_v2.update_quality_monitor` Update quality monitor (quality monitor v2) write - `databricks.quality_monitor_v2.delete_quality_monitor` Delete quality monitor (quality monitor v2) destructive **Aifunctions 4** - `databricks.ai_functions.ai_classify` Ai classify (ai functions) write - `databricks.ai_functions.ai_decide` Ai decide (ai functions) write - `databricks.ai_functions.ai_extract` Ai extract (ai functions) write - `databricks.ai_functions.ai_parse_document` Ai parse document (ai functions) write **Dataclassification 4** - `databricks.data_classification.create_catalog_config` Create catalog config (data classification) write - `databricks.data_classification.delete_catalog_config` Delete catalog config (data classification) destructive - `databricks.data_classification.get_catalog_config` Get catalog config (data classification) read - `databricks.data_classification.update_catalog_config` Update catalog config (data classification) write This index shows a sample of each area. All 1,371 actions, with their inputs and prices, are in the [catalogue](https://actionlake.com/console/catalogue), which a free sandbox key can search. Questions ## The Databricks MCP server, answered. **Is there a Databricks MCP server for Claude?** Yes. actionlake is an MCP gateway: a remote MCP server that gives Claude, or another MCP client, 1,371 Databricks actions across catalog, ML, settings and IAM, through the same eight tools it uses for every other connector. actionlake is independent; it is not made or endorsed by Databricks. **What do I need to connect Databricks?** Service principal application (client) ID and The service principal's OAuth secret (shown once when generated). Plus your account id and your tenant host. The free call account { op: "connections.requirements", service: "databricks" } lists exactly what to paste. **Can an AI agent delete or change things in Databricks?** Only if its key allows it, and only with an explicit confirmation. 231 Databricks actions are destructive, including delete (users), delete (groups) and delete (service principals). Each runs only with confirm: true. Under the human confirm policy, or an approval rule on the key, the run waits until a named person approves it. **Will an AI agent use up my Databricks API limit?** actionlake paces every call to the limits in the Databricks service file: 25 calls a second per connection. Repeated reads are answered from cache (539 Databricks actions can be), which never reaches Databricks. **What does a Databricks action cost?** One upstream call, the only unit actionlake bills. A list that pages bills one call per page (1,371 Databricks actions page). Cache hits, searches, dry runs and calls Databricks refused with 401, 403, 429 or a 5xx are free. The Free plan includes 100,000 calls a month with no card. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000. **Which Databricks API does it cover?** The Databricks REST API 2.0/2.1 as generated into databricks-sdk-py 0.156.0 at your Databricks instance's API base URL, which you enter when connecting. It is built from Databricks's documentation as read on 22 October 2026. 1,371 actions in 41 areas, each one in the signed service file (version 1.0.0). **Does actionlake keep my Databricks data, and where?** actionlake processes and stores data in Sydney, Australia; there is no US region yet. Reads may be cached for up to 5 minutes, and a write clears the reads it affects. Each run's request and response is kept, sealed, for your plan's trace window: 1 day on Free, 3 days on Starter, 7 days on Pro, 14 days on Team, 30 days on Business and 30 days on Enterprise. Your Databricks credential is sealed under your tenant's key, never logged and never returned. Your traffic is not used to train models. **Which AI clients can use it?** Claude Code, tested by us end to end, and the Claude app as a custom connector, tested by us with an existing key. Other MCP clients that support remote servers over Streamable HTTP (MCP revisions 2025-06-18 and 2025-11-25) and OAuth dynamic client registration, or a bearer header, should connect the same way; we have not tested them. Or plain REST, under the same action names. ## More connectors like Databricks Shares the most action names with - [Looker 12](https://actionlake.com/connectors/looker) - [Sigma Computing 11](https://actionlake.com/connectors/sigma-computing) - [Cribl 10](https://actionlake.com/connectors/cribl) - [Lightdash 10](https://actionlake.com/connectors/lightdash) - [MuleSoft Anypoint Platform 9](https://actionlake.com/connectors/mulesoft-anypoint-platform) - [Qlik Cloud 9](https://actionlake.com/connectors/qlik-cloud) Joins with, on shared entities - [Rubrik Security Cloud 3](https://actionlake.com/connectors/rubrik-security-cloud) - [Cloudflare 3](https://actionlake.com/connectors/cloudflare) - [Delinea Secret Server 3](https://actionlake.com/connectors/delinea-secret-server) - [authentik 3](https://actionlake.com/connectors/authentik) - [HashiCorp Vault 3](https://actionlake.com/connectors/hashicorp-vault) - [Spacelift 3](https://actionlake.com/connectors/spacelift) Browse - [All Observability & analytics connectors](https://actionlake.com/connectors/category/observability-and-analytics) - [All connectors](https://actionlake.com/connectors) - [How actionlake compares](https://actionlake.com/compare) actionlake is independent. It is not made or endorsed by Databricks; the name is the vendor's own. Counts are from the Databricks service file, version 1.0.0. ## Give your agent Databricks today. [Connect Databricks free](https://actionlake.com/console?start=sandbox) [See pricing](https://actionlake.com/pricing) Start free - 100,000 calls/month, no card. No signup. Hard caps on every key. Cached answers are free. --- Source: https://actionlake.com/connectors/databricks More: https://actionlake.com/llms.txt