Connect agents to tools

Connecting your agents to tools gives them practical capabilities beyond text generation, like searching documents, querying tables, calling external APIs, or running custom code.

The diagram shows the MCP route, which Databricks recommends for most integrations. MCP is one of several ways to connect agents to external services. Alongside MCP Services, you can call REST APIs directly through a Unity Catalog HTTP connection. Select Managed OAuth for per-user authentication, the Unity Catalog connections proxy to call APIs from agent code, or Unity Catalog function tools that wrap http_request(). The following table summarizes each way to connect, from Databricks data to external services:

Approach	Recommended use case
Managed MCP servers for Databricks data	Use this approach to query Databricks data and run governed functions with the ready-to-use Genie, Databricks AI Search, Databricks SQL, and Unity Catalog function MCP servers.
External MCP servers	Use this approach for services that publish an MCP server. It offers automatic tool discovery and governed access through Unity AI Gateway.
Managed OAuth	Use this approach for Google Drive or SharePoint integrations. Databricks manages the OAuth credentials, so no app registration is required.
Unity Catalog connections proxy	Use this approach to make direct REST API calls from agent code using the external service's own client SDK.
Unity Catalog function tools	Use this approach for SQL-based tool definitions that wrap the http_request() function.

Managed MCP servers for Databricks data

Databricks offers ready-to-use managed MCP servers that give agents governed access to data and functions in your workspace, with no server to build or host. Each server has a dedicated URL and OAuth scope, and Unity Catalog governs access:

• Genie: Query structured data in Genie Agents and Unity Catalog tables with natural language.
• Databricks AI Search (vector search): Search documents in vector search indexes.
• Databricks SQL: Run SQL queries against Unity Catalog tables.
• Unity Catalog functions: Call custom Python and SQL functions registered in Unity Catalog.

For server URLs, OAuth scopes, and the full catalog, see Databricks managed MCP servers. To call these servers from agent code, see Use MCP servers in agents.

External MCP servers

Connect your agents to external applications like Slack, Google Calendar, or any service with an API. Databricks offers several approaches depending on whether the external service has an MCP server, whether you need per-user authentication, or whether you prefer to call APIs directly from agent code. All approaches rely on a Unity Catalog HTTP connection, which provides secure, governed credential management and supports multiple authentication methods, including OAuth 2.0 user-to-machine (U2M) and machine-to-machine (M2M) authentication.

If the external service has an MCP server available, Databricks recommends registering it as an MCP Service in Unity Catalog. MCP Services provide automatic tool discovery, per-user authentication, and governance through Unity AI Gateway with grants, tool selection, and service policies.

• See Register an external MCP server to register an external MCP server as an MCP Service and govern it. For connection and authentication setup, see Step 1. Create a connection.
• See Use MCP servers in agents for code examples per agent framework (OpenAI Agents SDK, LangGraph, Model Serving).

For common software-as-a-service (SaaS) tools such as Slack, GitHub, and Google Drive, Databricks provides ready-to-use MCP Services with zero setup. See Databricks-provided MCP Services.

Managed OAuth

Databricks offers managed OAuth flows for select API tool providers. You don't need to register your own OAuth app or manage credentials. Databricks recommends Managed OAuth for development and testing. If production use cases require generating custom OAuth credentials, see the providers' documentation for more information.

The following integrations use OAuth credentials that Databricks manages and stores securely in the backend.

Provider	Configuration notes	Supported scopes	Description
Google Drive API	None	https://www.googleapis.com/auth/drive.readonly https://www.googleapis.com/auth/documents.readonly https://www.googleapis.com/auth/spreadsheets.readonly offline_access	Read-only access to Google Drive files, including Google Docs and Google Sheets.
Gmail API	None	https://www.googleapis.com/auth/gmail.readonly offline_access	Read-only access to Gmail messages, threads, drafts, and labels.
Google Calendar API	None	https://www.googleapis.com/auth/calendar.readonly offline_access	Read-only access to Google Calendar events, calendars, and free/busy information.
SharePoint API	None	https://graph.microsoft.com/User.Read https://graph.microsoft.com/User.ReadBasic.All https://graph.microsoft.com/Sites.Read.All https://graph.microsoft.com/Files.Read https://graph.microsoft.com/Files.Read.All https://graph.microsoft.com/Mail.Read https://graph.microsoft.com/Mail.ReadBasic https://graph.microsoft.com/Mail.Read.Shared https://graph.microsoft.com/MailboxFolder.Read https://graph.microsoft.com/MailboxItem.Read https://graph.microsoft.com/Calendars.Read https://graph.microsoft.com/Calendars.Read.Shared https://graph.microsoft.com/Chat.Read https://graph.microsoft.com/Chat.ReadBasic https://graph.microsoft.com/ChatMember.Read https://graph.microsoft.com/ChatMessage.Read https://graph.microsoft.com/Channel.ReadBasic.All https://graph.microsoft.com/ChannelMessage.Read.All https://graph.microsoft.com/OnlineMeetings.Read https://graph.microsoft.com/OnlineMeetingTranscript.Read.All https://graph.microsoft.com/OnlineMeetingAiInsight.Read https://graph.microsoft.com/OnlineMeetingArtifact.Read.All https://graph.microsoft.com/OnlineMeetingRecording.Read.All offline_access openid profile email	Read-only access via Microsoft Graph to SharePoint and OneDrive files, Outlook mail and calendar, and Microsoft Teams chats, channels, and meetings.

To set up managed OAuth, create an HTTP connection with the OAuth User to Machine Per User auth type and select your provider from the OAuth Provider drop-down menu. For detailed steps, see Create a connection to the external service.

The provider prompts each user to authorize on first use.

If needed, allowlist the following redirect URIs that managed OAuth uses:

Cloud	Redirect URI
AWS	https://oregon.cloud.databricks.com/api/2.0/http/oauth/redirect
Azure	https://westus.azuredatabricks.net/api/2.0/http/oauth/redirect
GCP	https://us-central1.gcp.databricks.com/api/2.0/http/oauth/redirect

For managed OAuth providers with a published MCP server, such as Glean, GitHub, Atlassian, and Slack, Databricks can manage the OAuth credentials when you register the server as an MCP Service. See Managed OAuth providers.

Unity Catalog connections proxy endpoint

Use the Unity Catalog connections proxy endpoint with the external service's own client SDK to call REST APIs directly from agent code. Point the SDK's base URL to the proxy endpoint and use your Databricks token as the API key. Databricks authenticates the request and automatically injects the external service's credentials from the Unity Catalog connection. Your code doesn't handle the external service's tokens directly.

Permissions required: USE CONNECTION on the connection object.

OpenAI

Use DatabricksOpenAI to route calls to external OpenAI through the Unity Catalog connections proxy. First, create a Unity Catalog HTTP connection using your OpenAI API key stored as a Databricks secret:

SQL

CREATE CONNECTION openai_connection TYPE HTTP
OPTIONS (
  host 'https://api.openai.com',
  base_path '/v1',
  bearer_token secret ('<secret-scope>', '<secret-key>')
);

Then install the databricks-openai package and use the proxy URL and workspace client in your agent code:

Bash

pip install databricks-openai

Python

from databricks_openai import DatabricksOpenAI
from databricks.sdk import WorkspaceClient

w = WorkspaceClient()

client = DatabricksOpenAI(
    workspace_client=w,
    base_url=f"{w.config.host}/api/2.0/unity-catalog/connections/openai_connection/proxy/",
)

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

Slack

Configure the Slack SDK to route through the Unity Catalog connections proxy. Create a Unity Catalog HTTP connection with host https://slack.com and base path /api, then use the proxy URL as the SDK base URL:

Python

from slack_sdk import WebClient
from databricks.sdk import WorkspaceClient

w = WorkspaceClient()

client = WebClient(
    token=w.config.authenticate()["Authorization"].split(" ")[1],
    base_url=f"{w.config.host}/api/2.0/unity-catalog/connections/slack_connection/proxy/",
)

result = client.chat_postMessage(channel="C123456", text="Hello from Databricks!")
print(result["message"]["text"])

Generic HTTP

For services without a dedicated SDK, use the requests library with the proxy URL directly:

Python

import requests
from databricks.sdk import WorkspaceClient

w = WorkspaceClient()

response = requests.post(
    f"{w.config.host}/api/2.0/unity-catalog/connections/my_connection/proxy/api/v1/resource",
    headers={
        **w.config.authenticate(),
        "Content-Type": "application/json",
    },
    json={"key": "value"},
)

For details on the proxy endpoint, supported authentication methods, and connection setup, see Forward requests through the HTTP connection proxy.

Unity Catalog function tools

note

Databricks recommends using MCP Services or the Unity Catalog connections proxy for new integrations. Unity Catalog function tools with http_request remain supported but are no longer the recommended approach.

You can create a Unity Catalog function that wraps http_request() to call external services. This approach is useful for SQL-based tool definitions, such as a function that posts a message to Slack. For the full walkthrough, including the SQL example and connection-type limitations, see Call external APIs with http_request (legacy).

Example notebooks

The following notebooks demonstrate creating AI agent tools that connect to Slack, OpenAI, and Azure AI Search.

Slack messaging agent tool

Open notebook in new tabOpen in Databricks

Microsoft Graph API agent tool

Open notebook in new tabOpen in Databricks

Azure AI Search agent tool

Open notebook in new tabOpen in Databricks

Additional resources

• Connect agents to third-party tools with MCP Services to register and govern external MCP servers.
• Create AI agent tools using Unity Catalog functions to build Unity Catalog function tools, including the legacy http_request approach.
• Integrate Unity Catalog tools with third party generative AI frameworks to use Unity Catalog tools with LangChain, LlamaIndex, OpenAI, and Anthropic.
• Connect to external HTTP services to configure HTTP connections and the connections proxy.