Feat/a2a expose usage metadata

[jvargassanchez-dot]

Summary

This PR introduces the functionality to expose usageMetadata from the Gemini API's streaming responses within the a2a-server in agent mode.

Details

Changes:

• Modified stream handling in a2a-server to extract usageMetadata.
• Includes the usageMetadata object in the final event sent to the client.
• Added comprehensive unit tests to validate the new logic.

Reasoning:

This change is required to enable token usage tracking and display in Gemini Code Assist, providing users with insights into context window utilization.

Related Issues

Fixes b/436499242 # Internal contribution for GCA feature

How to Validate

1. Run unit tests for the A2A server:
   bash npm test -w @google/gemini-cli-a2a-server
2. Verify that all 131 tests pass, confirming that both default policy loading and YOLO mode overrides are correctly handled.

Pre-Merge Checklist

• [ x] Updated relevant documentation and README (if needed)
• [ x] Added/updated tests (if needed)
• Noted breaking changes (if any)
• Validated on required platforms/methods:
  • MacOS
    • npm run
    • npx
    • Docker
    • Podman
    • Seatbelt
  • Windows
    • npm run
    • npx
    • Docker
  • [ x] Linux
    • [ x] npm run
    • npx
    • Docker

[gemini-code-assist]

Summary of Changes

Hello, 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 enables token usage tracking for the Gemini Code Assist feature by exposing usageMetadata from the agent's streaming responses. By capturing these metrics at the end of a task and including them in the final status update, the system can now provide users with visibility into their context window utilization.

Highlights

• Usage Metadata Extraction: Implemented logic to capture and store usageMetadata from Gemini API streaming responses within the Task class.
• Event Propagation: Updated the status update mechanism to include usageMetadata in the final event sent to the client.
• Testing: Added a new unit test case to verify that usageMetadata is correctly captured on Finished events and propagated in final status updates.

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 by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

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 pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize the 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 counterproductive. 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.

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. ↩

[gemini-code-assist]

Code Review

This pull request implements token usage tracking in the Task class of the a2a-server package. It adds a usageMetadata field to store token counts, captures this information from GeminiEventType.Finished events, and includes it in the final status update metadata. A corresponding test case was added to task.test.ts to ensure correct functionality. I have no feedback to provide.