Skip to content
Status

Google Sheets

Connect Google Sheets data through Google's hosted Sheets MCP server.

Use a Google Cloud project where you can enable services, grant project IAM roles, configure the Google Auth platform, and create OAuth credentials. You normally need Service Usage Admin or Owner to enable the APIs and Project IAM Admin to grant access. Each person who will connect needs access to the intended spreadsheets. Before setup, have the application or security owner configure prompt and response screening for malicious content or prompt injection.

Sign in to the Google Cloud console. In the console toolbar, use the resource selector to choose the project that will own this configuration. Keep the same project selected throughout setup.

Enable the Google Sheets APIs

  1. Open APIs & Services > API Library.
  2. In Search for APIs & Services, enter Google Sheets API.
  3. Open Google Sheets API.
  4. Click Enable. If the service is already enabled, leave it enabled.
  5. Reopen APIs & Services > API Library.
  6. In Search for APIs & Services, enter Google Sheets MCP API.
  7. Open Google Sheets MCP API.
  8. Click Enable. If the service is already enabled, leave it enabled.

Grant MCP Tool User access

  1. Go to IAM.
  2. Confirm that the same project is selected.
  3. Click Grant access.
  4. In New principals, enter the Google Account email address of a person who will connect through the Speakeasy AI Control Plane.
  5. Click Select a role.
  6. Search for MCP Tool User.
  7. Select MCP Tool User.
  8. Click Save.
  9. Repeat these steps for every person who will connect.

Google does not allow you to remove the OAuth consent screen after you configure it. Before continuing, confirm that you are working in the correct project.

  1. Open Google Auth platform > Branding.
  2. If Google Auth platform not configured yet appears, click Get Started.
  3. Under App Information, enter Sheets MCP Server in App name.
  4. In User support email, select a monitored email address. Obtain the correct value from the application or security owner if needed.
  5. Click Next.
  6. Under Audience, select Internal if every connecting user belongs to the project’s organization. Otherwise, select External.
  7. Click Next.
  8. Under Contact Information, enter a monitored Email address.
  9. Click Next.
  10. Under Finish, review the Google API Services User Data Policy with the application or security owner.
  11. If the owner approves the policy, select I agree to the Google API Services: User Data Policy.
  12. Click Continue.
  13. Click Create.

If the Google Auth platform was already configured, retain the approved Branding and Audience settings.

  1. Open Data Access.
  2. Click Add or Remove Scopes.
  3. Under Manually add scopes, paste these four URLs:
    • https://www.googleapis.com/auth/drive.readonly
    • https://www.googleapis.com/auth/drive.file
    • https://www.googleapis.com/auth/spreadsheets.readonly
    • https://www.googleapis.com/auth/spreadsheets
  4. Click Add to Table.
  5. Click Update.
  6. Click Save.
  7. If you selected an External audience in Testing, open Audience.
  8. Under Test users, click Add users.
  9. Enter each connecting user’s email address.
  10. Click Save.

An External audience in Testing supports up to 100 test users. Each authorization expires seven days after consent. When it expires, the user must complete browser authorization again.

Create the OAuth client

  1. Open Google Auth platform > Clients.
  2. Click Create client.
  3. In Application type, select Web application.
  4. In Name, enter a recognizable name such as Speakeasy AI Control Plane.
  5. Under Authorized redirect URIs, click + Add URI.
  6. In URIs, enter https://app.getgram.ai/mcp/remote_login_callback.

Before you click Create, prepare an approved secret store. The next dialog displays a client secret that can be copied only once.

  1. Click Create.

This opens OAuth 2.0 client created.

Copy the OAuth credentials

  1. In OAuth 2.0 client created, copy the Client ID to your approved secret store.
  2. Under Client secrets, copy the Client secret to the same store.

If you miss the one-time Client secret, delete it and create a new one before continuing.

Keep both values available, then connect your credentials in the Speakeasy AI Control Plane.

Add the server in Speakeasy

  1. In the Speakeasy AI Control Plane sidebar, under Connect, select Sources.
  2. Click Add Source.
  3. Choose Custom remote server.
  4. On Add a custom remote MCP server, paste https://sheetsmcp.googleapis.com/mcp/v1 into Remote MCP server URL.
  5. Click Add server.

This creates the hosted MCP server and opens its Overview page.

Connect your credentials

  1. From Overview, open Settings.
  2. Under Authentication, click Configure Manually or Use Discovered when offered.
  3. In Attach Remote Identity Provider, set Client Type to Manual.
  4. Confirm that Redirect URI matches https://app.getgram.ai/mcp/remote_login_callback from Create the OAuth client.
  5. In Client ID, paste the Client ID from Copy the OAuth credentials.
  6. In Client Secret (optional), paste the Client secret from the same step. Google requires this generated secret.
  7. In Scope (override), enter https://www.googleapis.com/auth/drive.readonly,https://www.googleapis.com/auth/drive.file,https://www.googleapis.com/auth/spreadsheets.readonly,https://www.googleapis.com/auth/spreadsheets.
  8. Click Attach Identity Provider.

At first connection, complete Google’s browser authorization with an account granted MCP Tool User and access to the intended spreadsheets.

This guide covers setup only. For anything beyond it — billing, tool behavior, limits — see Google’s Sheets MCP documentation at https://developers.google.com/workspace/sheets/api/guides/configure-mcp-server.