Databricks-provided MCP Services
Databricks provides ready-to-use MCP Services in the system.ai schema. These services include workspace tools and integrations with common SaaS applications, so agents can reach these tools without hosting or registering their own MCP server.
Each service is a Unity Catalog securable invoked through Unity Gateway. Built-in services ship with platform-managed tools and a built-in service policy, such as one to block write operations. You govern them with grants rather than with custom tool selection or policy functions.
Workspace tools
The DBSQL, web search, and sandbox MCP Services are in Beta.
MCP Service | Description |
|---|---|
Runs SQL on a SQL warehouse using the caller's Unity Catalog and warehouse permissions. | |
| Searches the public web and returns a synthesized answer with citations. See Web search on Databricks. |
| Runs Python, SQL, or shell code in an isolated environment. |
Connected applications
MCP Service | Connects to |
|---|---|
| Slack |
| GitHub |
| Jira and Confluence |
| Google Drive |
| Google Calendar |
| Gmail |
| Microsoft 365 (SharePoint, Outlook, and Teams) |
For Google Drive, Gmail, Google Calendar, and Microsoft 365, each user completes a one-time OAuth login before their first call. Open the service in Catalog Explorer and click Login.
Invoke a service
Address a service by its Unity Gateway URL, with its fully qualified Unity Catalog name in the path:
https://<workspace-hostname>/ai-gateway/mcp-services/system.ai.<service-name>
To call a service from agent code, see Use MCP tools in an agent.
Limitations
system.ai.web_searchis not available in workspaces with HIPAA/BAA compliance enabled. It is available in supported US regions or in workspaces that allow cross-Geo processing.
system.ai.sandboxdoes not have network egress.