Authentication

Agent Studio REST APIs support OAuth 2.0 Client Credentials Grant for authentication. This guide explains how to obtain an access token to authenticate your API requests. To view the full list of REST API endpoints, see the API Reference.

Prerequisites

Create an OAuth 2.0 Client

Follow the Authentication Guide to create an OAuth 2.0 Client in your Alation instance. Make sure to note down the client_id and client_secret as you will need them to obtain an access token.

REST API authentication

To authenticate with the REST API using OAuth 2.0 Client Credentials Grant, you need to obtain an access token by making a POST request to the token endpoint.

Curl

For a one-time request or testing purposes, you can use curl to obtain an access token. The access token expires after a certain period and you will need to automate the token refresh process.

curl --location '<base_url>/oauth/v2/token/' \
--header 'Content-Type: application/x-www-form-urlencoded' \
--data-urlencode 'grant_type=client_credentials' \
--data-urlencode 'client_id=<client_id>' \
--data-urlencode 'client_secret=<client_secret>'

Once the access token is obtained, include it in the Authorization header of your API requests as follows:

curl --location '<base_url>/ai/api/v1/chats/agent/default/catalog_search_agent/stream' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer <access_token>' \
--data '{
    "message": "Find me sales related tables"
}'

Python

We recommend using the Agent Studio SDK for interacting with the REST API. The SDK handles the authentication process for you, including obtaining and refreshing access tokens.

from alation_ai_agent_sdk import (
    AlationAIAgentSDK,
    ServiceAccountAuthParams,
)

BASE_URL = '<base_url>'
CLIENT_ID = '<client_id>'
CLIENT_SECRET = '<client_secret>'

sdk = AlationAIAgentSDK(
    base_url=BASE_URL,
    auth_method="service_account",
    auth_params=ServiceAccountAuthParams(
        client_id=CLIENT_ID,
        client_secret=CLIENT_SECRET,
    )
)

# Agents and tools can be called through the SDK
response = sdk.catalog_context_search_agent("Find me sales related tables")
# Or can be called through the API client
streamed_response = sdk.api.catalog_context_search_agent_stream("Find me sales related tables")

Without the SDK, you can use the requests library to obtain an access token and use it to call the API.

import requests

BASE_URL = '<base_url>'
CLIENT_ID = '<client_id>'
CLIENT_SECRET = '<client_secret>'

url = f'{BASE_URL}/oauth/v2/token/'
payload = {
  'grant_type': 'client_credentials',
  'client_id': CLIENT_ID,
  'client_secret': CLIENT_SECRET
}
response = requests.post(url, data=payload)
access_token = response.json().get('access_token')
print(access_token)

headers = {
    "Accept": "text/event-stream",
    "Content-Type": "application/json",
    "Authorization": f"Bearer {access_token}"
}

with requests.post(
    f'{BASE_URL}/ai/api/v1/chats/agent/default/catalog_search_agent/stream',
    headers=headers,
    json={"message": "Find me sales related tables"},
    stream=True
) as response:
    for line in response.iter_lines():
        if line:
            print(line.decode('utf-8'))