GitHub agent example

[ThomasCantuti]

Example of Google ADK agent using MCP protocol to connect with GitHub server

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[gemini-code-assist]

Summary of Changes

Hello @ThomasCantuti, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request adds a new example application that showcases how to build a conversational agent for GitHub. The agent utilizes Google's Agent Development Kit (ADK) and the Model Context Protocol (MCP) to interact with GitHub APIs, powered by a local Large Language Model. The primary goal is to provide a practical demonstration of integrating ADK with external services like GitHub using a flexible, tool-based approach.

Highlights

• New GitHub Agent Example: Introduces a new example project demonstrating a conversational AI agent capable of interacting with GitHub repositories using the Google Agent Development Kit (ADK) and the Model Context Protocol (MCP).
• Local LLM Integration: The agent is configured to connect with local Large Language Models (LLMs) via litellm and Ollama, specifically using a Qwen3 model, allowing for offline or self-hosted LLM operations.
• Dynamic Tooling for GitHub: Leverages ADK's MCPToolset to provide the agent with dynamic tools for querying GitHub APIs, enabling it to explore repository structures, analyze code, review issues, and more.
• Comprehensive Setup and Usage Guide: Includes a detailed README.md file that outlines prerequisites, installation steps, environment variable configuration (for GitHub Token), and instructions for running and interacting with the agent.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in issue comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[adk-bot]

Response from ADK Triaging Agent

Hello @ThomasCantuti, thank you for creating this PR!

To help us review your contribution, could you please address the following points from our contribution guidelines:

• Sign the Contributor License Agreement (CLA): It looks like the CLA check has failed. Please visit https://cla.developers.google.com/ to sign the agreement.
• Associate an Issue: For new features and examples like this one, please create an issue that describes the motivation and then link it to this PR.
• Add a Testing Plan: Please add a "Testing Plan" section to your PR description explaining how you've tested these changes and how a reviewer can verify them.

You can find more details in our contribution guidelines.

Thanks!

[gemini-code-assist]

Code Review

This pull request introduces a valuable example of a GitHub agent using the Google ADK and MCP protocol. The implementation is a great starting point. I've identified a few critical issues related to module imports and environment variable handling that will prevent the code from running as is. Additionally, I've included several suggestions to improve code quality, maintainability, and adherence to best practices, such as refining the authentication setup, improving configuration management, and making the code more robust. Addressing these points will make the example more reliable and easier for others to use.

[gemini-code-assist]

The authentication setup here is incorrect and confusing for a few reasons:

1. Incorrect Auth Scheme: You're using token_type='apikey' with name="GitHub-Token". However, the GitHub Copilot MCP API expects a bearer token in the Authorization header. The auth scheme declared to the model should match the actual auth mechanism. You should use token_type='oauth2Token'.

2. Redundant Header: The Authorization header is hardcoded in StreamableHTTPConnectionParams. The ADK's authentication framework is designed to handle this automatically via the auth_scheme and auth_credential you pass to the toolset. This redundancy is confusing and can lead to errors. You should remove the hardcoded header and rely on the framework's auth handling.

Here is a corrected version that addresses both issues.

    github_auth_scheme, github_auth_credential = token_to_scheme_credential(
        token_type='oauth2Token',
        credential_value=GITHUB_TOKEN
    )
    return MCPToolset(
        connection_params=StreamableHTTPConnectionParams(
            url="https://api.githubcopilot.com/mcp/",
        ),
        auth_credential=github_auth_credential,
        auth_scheme=github_auth_scheme,
        # tool_filter=["get_me", "list_pull_requests", "request_copilot_review", "merge_pull_request"]
    )