Fact-check statements with an ADK AI agent and Gemini model (original) (raw)

Coding level: Advanced
Duration: 30 minutes
Project type: Custom function

Overview

A fact-check custom function for Google Sheets to be used as a bound Apps Script project powered by a Vertex AIagent and Gemini model.

This sample demonstrates how you can use two powerful types of AI resources directly into your Google Sheets spreadsheets:

1. AI agents for sophisticated, multi-tool, multi-step reasoning capabilities using ADK agents deployed in the Vertex AI Agent Engine.
2. AI models for advanced understanding, generative, and summarization capabilities using Gemini models from Vertex AI.

Objectives

• Understand what the solution does.
• Understand how the solution is implemented.
• Deploy the Vertex AI agent.
• Set up the script.
• Run the script.

About this solution

The Sheets custom function is named FACT_CHECK and it operates as an end-to-end solution. It analyzes a statement, grounds its response using the latest web information, and returns the result in the format you need:

• Usage:
  • =FACT_CHECK("Your statement here") for a concise and summarized output.
  • =FACT_CHECK("Your statement here", "Your output formatting instructions here") for a specific output format.
• Reasoning: LLM Auditor ADK AI Agent (Python sample).
• Output formatting: Gemini model.

This solution requests Vertex AI REST APIs usingUrlFetchApp.

Architecture

The following diagram shows the architecture of the Google Workspace and Google Cloud resources used by the custom function.

Prerequisites

To use this sample, you need the following prerequisites:

• A Google Account (Google Workspace accounts might require administrator approval).
• A web browser with access to the internet.
• Prerequisites of the LLM Auditor ADK agent
  • Python 3.11+: For installation, follow instructions on the officialPython website.
  • Python Poetry: For installation, follow instructions on the officialPoetry website.
  • Google Cloud CLI: For installation, follow instructions on the officialGoogle Cloud website.

Prepare the environment

This section shows how to create and configure a Google Cloud project.

Create a Google Cloud project

Google Cloud console

1. In the Google Cloud console, go to Menu > IAM & Admin > Create a Project.
   Go to Create a Project
2. In the Project Name field, enter a descriptive name for your project.
   Optional: To edit the Project ID, click Edit. The project ID can't be changed after the project is created, so choose an ID that meets your needs for the lifetime of the project.
3. In the Location field, click Browse to display potential locations for your project. Then, click Select.
4. Click Create. The Google Cloud console navigates to the Dashboard page and your project is created within a few minutes.

gcloud CLI

In one of the following development environments, access the Google Cloud CLI (gcloud):

• Cloud Shell: To use an online terminal with the gcloud CLI already set up, activate Cloud Shell.
  Activate Cloud Shell
• Local Shell: To use a local development environment,install andinitialize the gcloud CLI.
  To create a Cloud project, use the gcloud projects create command:

gcloud projects create PROJECT_ID  

Replace PROJECT_ID by setting the ID for the project you want to create.

Enable billing for the Cloud project

Google Cloud console

1. In the Google Cloud console, go to Billing. ClickMenu > Billing > My Projects.
   Go to Billing for My Projects
2. In Select an organization, choose the organization associated with your Google Cloud project.
3. In the project row, open the Actions menu (), click Change billing, and choose the Cloud Billing account.
4. Click Set account.

gcloud CLI

1. To list available billing accounts, run:

gcloud billing accounts list  

1. Link a billing account with a Google Cloud project:

gcloud billing projects link PROJECT_ID --billing-account=BILLING_ACCOUNT_ID  

Replace the following:

• PROJECT_ID is the Project ID for the Cloud project for which you want to enable billing.
• BILLING_ACCOUNT_ID is the billing account ID to link with the Google Cloud project.

Enable the Vertex AI API

Google Cloud console

1. In the Google Cloud console, enable the Vertex AI, and Cloud Resource Manager APIs.
   Enable the APIs
2. Confirm that you're enabling the Vertex AI API in the correct Cloud project, then click Next.
3. Confirm that you're enabling the correct API, then click Enable.

gcloud CLI

1. If necessary, set the current Cloud project to the one you created with the gcloud config set project command:

gcloud config set project PROJECT_ID  

Replace PROJECT_ID with the Project ID of the Cloud project you created. 2. Enable the Vertex AI API with the gcloud services enable command:

gcloud services enable aiplatform.googleapis.com  

Create a service account in Google Cloud console

Create a new service account with the role Vertex AI User by following these steps:

Google Cloud console

1. In the Google Cloud console, go to Menu > IAM & Admin > Service Accounts.
   Go to Service Accounts
2. Click Create service account.
3. Fill in the service account details, then click Create and continue.
4. Optional: Assign roles to your service account to grant access to your Google Cloud project's resources. For more details, refer to Granting, changing, and revoking access to resources.
5. Click Continue.
6. Optional: Enter users or groups that can manage and perform actions with this service account. For more details, refer to Managing service account impersonation.
7. Click Done. Make a note of the email address for the service account.

gcloud CLI

1. Create the service account:

gcloud iam service-accounts create SERVICE_ACCOUNT_NAME \  
  --display-name="SERVICE_ACCOUNT_NAME"  

1. Optional: Assign roles to your service account to grant access to your Google Cloud project's resources. For more details, refer to Granting, changing, and revoking access to resources.

The service account appears on the service account page. Next, create a private key for the service account.

Create a private key

To create and download a private key for the service account, follow these steps:

1. In the Google Cloud console, go to Menu > IAM & Admin > Service Accounts.
   Go to Service Accounts
2. Select your service account.
3. Click Keys > Add key > Create new key.
4. Select JSON, then click Create.
   Your new public/private key pair is generated and downloaded to your machine as a new file. Save the downloaded JSON file as credentials.json in your working directory. This file is the only copy of this key. For information about how to store your key securely, seeManaging service account keys.
5. Click Close.

For more information about service accounts, seeservice accountsin the Google Cloud IAM documentation.

Deploy the LLM Auditor ADK AI Agent

1. If not done already, authenticate with your Google Cloud account and configure Google Cloud CLI to use your Google Cloud project.
   gcloud auth application-default login
gcloud config set project PROJECT_ID
gcloud auth application-default set-quota-project PROJECT_ID
   Replace PROJECT_ID with the ID of the Cloud project you created.
2. Download this GitHub repository
   Download
3. In your preferred local development environment, extract the downloaded archive file and open the adk-samples/python/agents/llm-auditor directory.
   unzip adk-samples-main.zip
cd adk-samples-main/python/agents/llm-auditor
4. Create a new Cloud Storage bucket dedicated to the ADK agent.

gcloud storage buckets create gs://CLOUD_STORAGE_BUCKET_NAME --project=PROJECT_ID --location=PROJECT_LOCATION  

Replace the following:

1. CLOUD_STORAGE_BUCKET_NAME with a unique bucket name you want to use.
2. PROJECT_ID with the ID of the Cloud project you created.
3. PROJECT_LOCATION with the location of the Cloud project you created.
4. Set the following environment variables:
   export GOOGLE_GENAI_USE_VERTEXAI=true
export GOOGLE_CLOUD_PROJECT=PROJECT_ID
export GOOGLE_CLOUD_LOCATION=PROJECT_LOCATION
export GOOGLE_CLOUD_STORAGE_BUCKET=CLOUD_STORAGE_BUCKET_NAME
   Replace the following:
   1. CLOUD_STORAGE_BUCKET_NAME with the name of the bucket you created.
   2. PROJECT_ID with the ID of the Cloud project you created.
   3. PROJECT_LOCATION with the location of the Cloud project you created.
5. Install and deploy ADK agent from virtual environment.
   python3 -m venv myenv
source myenv/bin/activate
poetry install --with deployment
python3 deployment/deploy.py --create
6. Retrieve the agent ID, you'll need it later to configure the custom function.

python3 deployment/deploy.py --list  

Review the sample code

Optionally, before creating the new spreadsheet, take a moment to review and familiarize yourself with the sample code hosted on GitHub.

View on GitHub

Create and configure in a new spreadsheet

1. Click the following button and make a complete copy of the sample Google Sheets spreadsheet including its container-bound Apps Script project.
   Copy Google Sheets speadsheet
2. In the newly created spreadsheet, go to Extensions > Apps Script.
3. In the Apps Script project, go to Project Settings, clickEdit script properties, then click Add script property to add the following script properties:
   1. LOCATION with the location of the Google Cloud project created in previous steps such as us-central1.
   2. GEMINI_MODEL_ID with the Gemini model you want to use such asgemini-2.5-flash-lite.
   3. REASONING_ENGINE_ID with the ID of the LLM Auditor ADK Agent deployed in previous steps such as 1234567890.
   4. SERVICE_ACCOUNT_KEY with the JSON key from the service account downloaded in previous steps such as { ... }.
4. Click Save script properties

Test the custom function

1. Go to the newly created spreadsheet.
2. Change statements in column A.
3. Formulas in column B execute and then display fact-check results.

Clean up

To avoid incurring charges to your Google Cloud account for the resources used in this tutorial, we recommend that you delete the Cloud project.

1. In the Google Cloud console, go to the Manage resources page. ClickMenu > IAM & Admin > Manage Resources.
   Go to Resource Manager
2. In the project list, select the project you want to delete and then clickDelete .
3. In the dialog, type the project ID and then click Shut down to delete the project.

Next steps

• Plan travels with an AI agent accessible across Google Workspace
• Custom functions in Google Sheets
• Extending Google Sheets