Google_GenerativeAI.Auth 2.0.2

Google_GenerativeAI.Auth

Google_GenerativeAI.Auth is a helper library for the Unofficial C# Google Generative AI SDK (Google Gemini) that simplifies authentication using OAuth and Service Accounts. It provides concrete implementations for the IGoogleAuthenticator interface, making it easy to integrate authentication into your applications using the SDK. This library handles the complexities of obtaining access tokens, so you can focus on building generative AI applications.

Features

• Service Account Authentication: Authenticate using Google Service Accounts, either with a JSON key file or with a key and password.
• IGoogleAuthenticator Implementation: Seamlessly integrates with the Google Generative AI SDK by implementing the IGoogleAuthenticator interface.
• Easy-to-Use API: Provides simple methods for creating authenticators and retrieving access tokens.
• Asynchronous Support: Uses async/await for non-blocking, asynchronous operations.

Installation

Install the Google_GenerativeAI.Auth NuGet package:

dotnet add package Google_GenerativeAI.Auth

You will also need to install the main Google Generative AI SDK:

dotnet add package Google_GenerativeAI  

Usage

The Google_GenerativeAI.Auth library provides two main ways to authenticate with a service account:

1. Using a JSON Key File (Recommended):
   Uses the path to a service account JSON key file. This is the preferred method for most use cases.

2. Using Key and Password:
   Requires the service account email, key, and password.

1. Service Account Authentication (JSON Key File)

This is the recommended approach. Download a JSON key file for your service account from the Google Cloud Console. Keep this file secure!

// Assuming 'jsonFilePath' is the path to your service account JSON key file.
var authenticator = new GoogleServiceAccountAuthenticator(jsonFilePath);
var vertexAi = new VertexAIModel(authenticator: authenticator);

Important: jsonFilePath should be the full path to your JSON key file. Do not hardcode this path. Instead, retrieve it from an environment variable or a secure configuration store:

2. Service Account Authentication (Key and Password)

This method is less common and requires careful handling of the credentials.

// Assuming 'email', 'key', and 'password' are your service account credentials.
var authenticator = new GoogleServiceAccountAuthenticator(email, key, password);
var vertexAi = new VertexAIModel(authenticator: authenticator);

Important: email, key, and password should be your actual service account credentials. Do not hardcode these values. Use environment variables or a secure configuration system:

2. OAuth Authentication with Client Secret and Client Id

// Assuming 'credentialFile' your client_secret.json file.
var authenticator = new GoogleOAuthAuthenticator(credentialFile);
var vertexAi = new VertexAIModel(authenticator: authenticator);

Getting the Access Token Directly

You can get the access token directly from the IGoogleAuthenticator if needed:

var token = await authenticator.GetAccessTokenAsync();
Console.WriteLine(token.AccessToken); // Use the token as needed.

Important Security Considerations

1. NEVER hardcode credentials directly in your source code. Use environment variables, configuration files, or a secrets management service (like Azure Key Vault or AWS Secrets Manager).
2. Protect your service account JSON key file. Store it securely and do not commit it to version control. Treat it like a password. Add it to your .gitignore file.
3. Grant the least privilege necessary to your service account. Only give it the permissions required to access the Google Generative AI APIs.

Dependencies

• Google Generative AI SDK (C#) (This is a peer dependency.)
• Google.Apis.Auth

Contributing

Contributions are welcome! Please open an issue to discuss proposed changes or create a pull request.

License

MIT License