Authenticate using 2-legged OAuth with auth manager

To let your agents authenticate to external tools like ServiceNow or Salesforce using their own authority, configure outbound authentication using 2-legged OAuth (Client Credentials) auth providers in Agent Identity auth manager.

2-legged OAuth auth providers manage credentials and tokens for you. This removes the need to write custom code to handle authentication flows.

2-legged OAuth workflow

2-legged OAuth auth providers use the agent's identity and don't require user consent. Google manages the storage of the client credentials. When you use the Agent Development Kit (ADK), it automatically retrieves and injects the resulting access tokens into the tool invocation headers.

Before you begin

Verify that you have chosen the correct authentication method.
Enable the Agent Identity Connector API.
Roles required to enable APIs
To enable APIs, you need the Service Usage Admin IAM role (roles/serviceusage.serviceUsageAdmin), which contains the serviceusage.services.enable permission. Learn how to grant roles.
Enable the API
Create and deploy an agent.
Obtain the client ID and client secret from the third-party application that you want to connect to.
Verify that you have the roles required to complete this task.

Required roles

To get the permissions that you need to create and use a 2-legged Agent Identity auth provider, ask your administrator to grant you the following IAM roles on the project:

To create auth providers:
- IAM Connector Admin (roles/iamconnectors.admin)
- IAM Connector Editor (roles/iamconnectors.editor)
To use auth providers:
- IAM Connector User (roles/iamconnectors.user)
- Agent Default Access (roles/aiplatform.agentDefaultAccess)
- Agent Context Editor (roles/aiplatform.agentContextEditor)
- Vertex AI User (roles/aiplatform.user)
- Service Usage Consumer (roles/serviceusage.serviceUsageConsumer)

For more information about granting roles, see Manage access to projects, folders, and organizations.

These predefined roles contain the permissions required to create and use a 2-legged Agent Identity auth provider. To see the exact permissions that are required, expand the Required permissions section:

Required permissions

The following permissions are required to create and use a 2-legged Agent Identity auth provider:

To create auth providers: iamconnectors.connectors.create
To use auth providers:
- iamconnectors.connectors.retrieveCredentials
- aiplatform.endpoints.predict
- aiplatform.sessions.create

You might also be able to get these permissions with custom roles or other predefined roles.

Create a 2-legged auth provider

Create an auth provider to define the configuration and credentials for third-party applications.

To create a 2-legged auth provider, use the Google Cloud console or the Google Cloud CLI.

Console

In the Google Cloud console, go to the Agent Registry page.
Go to Agent Registry
Click the name of the agent that you want to create an auth provider for.
Click Identity.
In the Auth Providers section, click Add auth provider.
In the Add auth provider pane, enter a name and description.

The name can contain only lowercase letters, numbers, or hyphens, cannot end with a hyphen, and must start with a lowercase letter.
From the OAuth Type list, select OAuth (2 legged) .
Click Create and continue.
To grant your agent identity permission to use the auth provider, click Grant access.
This automatically assigns the Connector User (roles/iamconnectors.user) role to the agent identity on the auth provider resource.
In the Auth provider credentials section, enter the following information:
- Client ID
- Client Secret
- Token URL
Click Add provider config.

The newly created auth provider appears in the Auth Providers list.

Google Cloud CLI

Create the auth provider:

gcloud alpha agent-identity connectors create AUTH_PROVIDER_NAME \
    --location="LOCATION" \
    --two-legged-oauth-client-id="CLIENT_ID" \
    --two-legged-oauth-client-secret="CLIENT_SECRET" \
    --two-legged-oauth-token-endpoint="TOKEN_ENDPOINT"

To grant your agent identity permission to use auth providers, update the IAM allow policy for either the project or the specific auth provider, and grant the Connector User (roles/iamconnectors.user) role to the agent principal.

Agent Identity is based on the industry-standard SPIFFE ID format. In IAM allow policies, agent identities are referred to using principal identifiers.
Project-level (gcloud)

Granting the role at the project level allows the agent to use any auth provider in that project.
- To grant a single agent access to auth providers in a project, run the following command:
  gcloud projects add-iam-policy-binding PROJECT_ID \ --role='roles/iamconnectors.user' \ --member="principal://agents.global.org-ORGANIZATION_ID.system.id.goog/resources/aiplatform/projects/PROJECT_NUMBER/locations/LOCATION/reasoningEngines/ENGINE_ID"
- To grant all agents in a project access to auth providers, run the following command:
  gcloud projects add-iam-policy-binding PROJECT_ID \ --role='roles/iamconnectors.user' \ --member="principalSet://agents.global.org-ORGANIZATION_ID.system.id.goog/attribute.platformContainer/aiplatform/projects/PROJECT_NUMBER"
Connector-level (curl)

To grant a single agent access to a specific auth provider, use the setIamPolicy API. This command overwrites any existing allow policy on the resource.
```
curl -X POST \
    -H "Authorization: Bearer $(gcloud auth print-access-token)" \
    -H "Content-Type: application/json" \
    -d '{
  "policy": {
    "bindings": [
      {
        "role": "roles/iamconnectors.user",
        "members": ["principal://agents.global.org-ORGANIZATION_ID.system.id.goog/resources/aiplatform/projects/PROJECT_NUMBER/locations/LOCATION/reasoningEngines/ENGINE_ID"]
      }
    ]
  }
}' \
    "https://iamconnectors.googleapis.com/v1alpha/projects/PROJECT_ID/locations/LOCATION/connectors/AUTH_PROVIDER_NAME:setIamPolicy"
```
Replace the following:
- PROJECT_ID: Your Google Cloud project ID.
- AUTH_PROVIDER_NAME: The name of the auth provider.
- ORGANIZATION_ID: Your Google Cloud organization ID.
- PROJECT_NUMBER: Your Google Cloud project number.
- LOCATION: The location for your agent (for example, us-central1).
- ENGINE_ID: The ID of your reasoning engine.

Authenticate in your agent code

To authenticate your agent, you can use the ADK.

ADK

Reference the auth provider in your agent's code by using the MCP toolset in the ADK.

from google.adk.agents.llm_agent import LlmAgent
from google.adk.auth.credential_manager import CredentialManager
from google.adk.integrations.agent_identity import GcpAuthProvider, GcpAuthProviderScheme
from google.adk.tools.mcp_tool.mcp_session_manager import StreamableHTTPConnectionParams
from google.adk.tools.mcp_tool.mcp_toolset import McpToolset
from google.adk.auth.auth_tool import AuthConfig

# Register the Google Cloud Auth Provider so the CredentialManager can use it.
CredentialManager.register_auth_provider(GcpAuthProvider())

# Create the Google Cloud Auth Provider scheme using the auth provider's full resource name.
auth_scheme = GcpAuthProviderScheme(
    name="projects/PROJECT_ID/locations/LOCATION/connectors/AUTH_PROVIDER_NAME"
)

# Configure an MCP tool with the authentication scheme.
toolset = McpToolset(
    connection_params=StreamableHTTPConnectionParams(url="https://YOUR_MCP_SERVER_URL"),
    auth_scheme=auth_scheme,
)

# Initialize the agent with the authenticated tools.
agent = LlmAgent(
    name="YOUR_AGENT_NAME",
    model="gemini-2.0-flash",
    instruction="YOUR_AGENT_INSTRUCTIONS",
    tools=[toolset],
)

ADK

Reference the auth provider in your agent's code using an authenticated function tool in the ADK.

import httpx
from google.adk.agents.llm_agent import LlmAgent
from google.adk.auth.credential_manager import CredentialManager
from google.adk.integrations.agent_identity import GcpAuthProvider
from google.adk.integrations.agent_identity import GcpAuthProviderScheme
from google.adk.apps import App
from google.adk.auth.auth_credential import AuthCredential
from google.adk.auth.auth_tool import AuthConfig
from google.adk.tools.authenticated_function_tool import AuthenticatedFunctionTool
from vertexai import agent_engines

# First, register Google Cloud auth provider
CredentialManager.register_auth_provider(GcpAuthProvider())

# Create Auth Config
spotify_auth_config = AuthConfig(
    auth_scheme=GcpAuthProviderScheme(
        name="projects/PROJECT_ID/locations/LOCATION/connectors/AUTH_PROVIDER_NAME"
    )
)

# Use the Auth Config in Authenticated Function Tool
spotify_search_track_tool = AuthenticatedFunctionTool(
    func=spotify_search_track, auth_config=spotify_auth_config
)

# Sample function tool
async def spotify_search_track(credential: AuthCredential, query: str) -> str | list:
    token = None
    if credential.http and credential.http.credentials:
        token = credential.http.credentials.token

    if not token:
        return "Error: No authentication token available."

    async with httpx.AsyncClient() as client:
        response = await client.get(
            "https://api.spotify.com/v1/search",
            headers={"Authorization": f"Bearer {token}"},
            params={"q": query, "type": "track", "limit": 1},
        )
        # Add your own logic here

agent = LlmAgent(
    name="YOUR_AGENT_NAME",
    model="YOUR_MODEL_NAME",
    instruction="YOUR_AGENT_INSTRUCTIONS",
    tools=[spotify_search_track_tool],
)

app = App(
    name="YOUR_APP_NAME",
    root_agent=agent,
)

vertex_app = agent_engines.AdkApp(app_name=app)

ADK

Reference the auth provider in your agent's code using the Agent Registry MCP toolset in the ADK.

  from google.adk.agents.llm_agent import LlmAgent
from google.adk.auth.credential_manager import CredentialManager
from google.adk.integrations.agent_identity import GcpAuthProvider
from google.adk.integrations.agent_identity import GcpAuthProviderScheme
from google.adk.tools.mcp_tool.mcp_session_manager import StreamableHTTPConnectionParams
from google.adk.tools.mcp_tool.mcp_toolset import McpToolset
from google.adk.auth.auth_tool import AuthConfig
from google.adk.integrations.agent_registry import AgentRegistry

# First, register Google Cloud auth provider
CredentialManager.register_auth_provider(GcpAuthProvider())

# Create Google Cloud auth provider scheme by providing Auth Provider full resource name
auth_scheme = GcpAuthProviderScheme(
    name="projects/PROJECT_ID/locations/LOCATION/connectors/AUTH_PROVIDER_NAME"
)

# Set Agent Registry
registry = AgentRegistry(project_id="PROJECT_ID", location="global")

toolset = registry.get_mcp_toolset(mcp_server_name="projects/PROJECT_ID/locations/global/mcpServers/agentregistry-00000000-0000-0000-0000-000000000000", auth_scheme=auth_scheme)

# Example MCP tool
toolset = McpToolset(
    connection_params=StreamableHTTPConnectionParams(url="MCP_URL"),
    auth_scheme=auth_scheme,
)

agent = LlmAgent(
    name="YOUR_AGENT_NAME",
    model="YOUR_MODEL_NAME",
    instruction="YOUR_AGENT_INSTRUCTIONS",
    tools=[toolset],
)

Install dependencies for local testing

To test your agent locally in a virtual environment, install the following necessary dependencies:

Create and activate a virtual environment:

python3 -m venv env
source env/bin/activate

Install the required packages:

pip install google-cloud-aiplatform[agent_engines,adk] google-adk[agent-identity]

Deploy the agent

When you deploy your agent to Google Cloud, ensure that Agent Identity is enabled.

If you're deploying to Agent Runtime, use the identity_type=AGENT_IDENTITY flag:

import vertexai
from vertexai import types
from vertexai.agent_engines import AdkApp

# Initialize the Vertex AI client with v1beta1 API for Agent Identity support
client = vertexai.Client(
    project="PROJECT_ID",
    location="LOCATION",
    http_options=dict(api_version="v1beta1")
)

# Use the proper wrapper class for your Agent Framework (e.g., AdkApp)
app = AdkApp(agent=agent)

# Deploy the agent with Agent Identity enabled
remote_app = client.agent_engines.create(
    agent=app,
    config={
        "identity_type": types.IdentityType.AGENT_IDENTITY,
        "requirements": ["google-cloud-aiplatform[agent_engines,adk]", "google-adk[agent-identity]"],
    },
)