// SCAN REPORT

Databricks MCP Server

38 tools. 22 can modify or destroy data without limits.

5 destructive tools with no built-in limits. Policy required.

Last updated: 11 Jun 2026

22 can modify or destroy data

16 read-only

38 tools total

Community server · catalogue entry verified 11/06/2026

How to control Databricks MCP Server ↓

TOOL MAP

What Databricks MCP Server exposes to your agents

Read (16) Write / Execute (17) Destructive / Financial (5)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous Databricks MCP Server tools

Destructive · High dbfs_delete This tool permanently removes data from DBFS without possibility of recovery through normal means. Destructive · High delete_job Deleting a job is an irreversible operation that removes a defined workflow from the Databricks workspace. Destructive · High delete_workspace_object This tool irreversibly deletes data (notebooks or directories) from a Databricks workspace. Destructive · High uninstall_library Uninstalling libraries from a cluster is an irreversible removal action that could break running jobs, notebooks, or workflows depending on those libraries.

22 of Databricks MCP Server's 38 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control Databricks MCP Server

PolicyLayer is an MCP gateway — it sits between your AI agents and Databricks MCP Server, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations

{
  "dbfs_delete": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

{
  "pull_repo": {
    "limits": [
      {
        "counter": "pull_repo_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "get_cluster": {
    "limits": [
      {
        "counter": "get_cluster_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

Create a free account and register Databricks MCP Server — nothing to install.
Add these rules — paste them, or build them visually. Tune the limits to your setup.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

ENFORCE POLICY ON DATABRICKS →

Free to start. No card required.

FULL CATALOGUE

All 38 Databricks MCP Server tools

DESTRUCTIVE 5 tools

Destructive dbfs_delete Delete a DBFS path Destructive delete_job Delete a Databricks job Destructive delete_workspace_object Delete a workspace notebook or directory Destructive uninstall_library Uninstall libraries from a cluster Destructive terminate_cluster Terminate a cluster

EXECUTE 6 tools

Execute cancel_run Cancel a job run Execute execute_sql Execute a SQL statement Execute run_job Trigger a job run Execute run_notebook Submit a one-time notebook run Execute start_cluster Start a terminated cluster Execute sync_repo_and_run_notebook Pull a repo and run a notebook

WRITE 11 tools

Write pull_repo Pull latest commit for a repo Write create_catalog Create a Unity Catalog catalog Write create_cluster Create a new Databricks cluster Write create_job Create a Databricks job Write create_repo Create or clone a repo Write create_schema Create a schema in Unity Catalog Write create_table Create a table via SQL Write dbfs_put Upload small content to DBFS Write import_notebook Import a notebook Write install_library Install libraries on a cluster Write update_repo Update repo branch or tag

READ 16 tools

Read get_cluster Get information about a cluster Read get_run_status Get status for a job run Read get_table_lineage Get Unity Catalog table lineage Read get_workspace_file_content Retrieve workspace file content Read get_workspace_file_info Retrieve workspace metadata Read list_catalogs List Unity Catalog catalogs Read list_cluster_libraries List libraries for a cluster Read list_clusters List all Databricks clusters Read list_files List DBFS files for a path Read list_job_runs List recent job runs Read list_jobs List Databricks jobs Read list_notebooks List notebooks in a directory Read list_repos List repos in the workspace Read list_schemas List schemas in a catalog Read list_tables List tables for a schema Read export_notebook Export a notebook

EXPLORE

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

FAQ

Questions about Databricks MCP Server

Can an AI agent delete data through the Databricks MCP Server MCP server? +

Yes. The Databricks MCP Server server exposes 5 destructive tools including dbfs_delete, delete_job, delete_workspace_object. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through Databricks MCP Server? +

The Databricks MCP Server server has 11 write tools including pull_repo, create_catalog, create_cluster. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach Databricks MCP Server.

How many tools does the Databricks MCP Server MCP server expose? +

38 tools across 4 categories: Destructive, Execute, Read, Write. 16 are read-only. 22 can modify, create, or delete data.

How do I enforce a policy on Databricks MCP Server? +

Register the Databricks MCP Server MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every Databricks MCP Server tool call.

Deterministic rules across all 38 Databricks MCP Server tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON DATABRICKS →

Free to start. No card required.

38 Databricks MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.