Composio powers 1000+ toolkits, tool search, context management, authentication, and a sandboxed workbench to help you build AI agents that turn intent into action.

next
md 6.8 KB

How do I set up custom Google OAuth credentials for Google BigQuery?

Use your own Google OAuth app for production BigQuery connections. Composio's managed OAuth app is still waiting on Google verification for the BigQuery scope, so Google can block it (see the next question).

The Google account you connect needs IAM access to the BigQuery projects and datasets your workflow uses. For basic exploration, Google offers a limited BigQuery sandbox that doesn't need a billing account.

Create the Google OAuth app

  1. In Google Cloud Console, select the project that will own the OAuth client, or create one.

Google Cloud console project picker

  1. Open APIs & Services → Library, search for BigQuery API (bigquery.googleapis.com), and select Enable. If the page shows Manage, the API is already enabled.
  2. Open Google Auth Platform → Branding. If Google shows Get started, enter the app name, user support email, and developer contact email, then save.

Google Auth Platform Branding page with app name and user support email

  1. Open Audience. Choose Internal to limit the app to users in your Google Workspace organization, or External for consumer accounts and users outside it. While an external app's publishing status is Testing, only accounts listed under Test users can connect, so add every account you'll test with.

Google Auth Platform Audience page with Testing status, External user type, and Test users

  1. Open Data Access → Add or remove scopes and add the scopes listed under Scopes below. If a scope isn't in the list, paste its full URL under Manually add scopes and select Add to table. Select Update, then Save.
  2. Open Clients → Create client, choose Web application, and give it a name.

Create OAuth client ID form with Web application selected

  1. Under Authorized redirect URIs, add https://backend.composio.dev/api/v1/auth-apps/add. It must exactly match the oauth_redirect_uri field on your Composio auth config.

OAuth client form with the Composio redirect URI under Authorized redirect URIs

  1. Select Create. Copy the Client ID and Client secret. Google shows the full secret only once, so save it before you close the dialog.

Google OAuth web client details with Client ID and Client secret

Create the auth config in Composio

  1. In the Composio dashboard, open Auth Configs → Create and select Google BigQuery.
  2. Choose OAuth 2.0 and Your Own Credentials, then paste your Client ID and Client secret. Leave Redirect URI unchanged, keep the scopes in line with what you added in Google, and select Create Auth Config.
  3. Open the auth config, select Connect Account, and sign in with the Google account you want to connect. Review the requested permissions before you continue.

Google consent screen asking to view your data in Google BigQuery

  1. Open Accounts on the auth config and confirm the connected account shows Active. The auth config's own Enabled label doesn't confirm a connection.

Composio auth config panel with the Google BigQuery connected account shown as Active

For the general flow, see Custom auth configs.

Scopes

The default auth config requests https://www.googleapis.com/auth/cloud-platform, https://www.googleapis.com/auth/userinfo.email. Composio uses userinfo.email to identify the connected account, so keep it. To list projects with less access, you can use https://www.googleapis.com/auth/bigquery.readonly with userinfo.email instead. Add job or data-write scopes only if your workflow runs queries or loads data. See Google's projects.list reference for the required permissions.

Short scope names on this page leave out the https://www.googleapis.com/auth/ prefix. When you paste a scope under Manually add scopes, use the full URL, for example https://www.googleapis.com/auth/userinfo.email. Scopes already written as a full URL, such as https://mail.google.com/, go in as shown.

IAM, billing, and missing projects

OAuth scopes and Google Cloud IAM both apply. If a project or dataset is missing after you connect, check the signed-in user's IAM roles. Review BigQuery pricing and job permissions before you run queries or load data.

Use a service account instead

BigQuery also supports the Google Service Account auth scheme, which skips the user consent screen. Create an auth config with that scheme, then paste the service account's credentials JSON in the credentials_json field when you connect the account.

Publish and verify the app

While an external app is in Testing, Google shows an unverified-app notice and refresh tokens expire after seven days, so users have to reconnect. Before you invite production users, select Publish app on the Audience page. A public app that requests sensitive or restricted scopes also needs Google's OAuth verification, which you start from Verification Center. Publishing doesn't complete verification on its own. Internal apps and apps in Testing can be exempt. See Google's verification requirements.

If you change the requested scopes later, reconnect the account so the user grants the new access.

Why do I see "This app is blocked" when connecting Google BigQuery?

Google can show "This app is blocked" during the BigQuery OAuth flow when the OAuth app requests Google scopes that are not approved or verified for that app.

For managed Google BigQuery OAuth, this can happen because Composio's managed OAuth app does not yet have the required BigQuery scope approved. That scope is still under verification, which is why Google can block the connection.

To unblock production usage, use your own verified Google OAuth app in a custom auth config with the BigQuery scopes your workflow needs, then reconnect the account. If your setup can use service-account auth for BigQuery, that can also avoid the user-consent OAuth screen.