feat: Implement agent handoff tracking to skip after_agent_callback #4255
Conversation
Summary of ChangesHello @jayy-77, 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 enhances the agent orchestration within the ADK by introducing explicit tracking for agent handoffs. By adding an Highlights
🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console. Using Gemini Code AssistThe 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
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 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
|
adk-bot
added
the
core
|
Hello @jayy-77, thank you for creating this PR! Could you please provide a testing plan for your changes? This will help reviewers to review your PR more efficiently. Response from ADK Triaging Agent |
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces a feature to skip the after_agent_callback when an agent hands off control to another. The implementation correctly adds a flag to the invocation context and sets it during handoff. However, I've identified a significant logical issue where the receiving agent's after_agent_callback would also be skipped, which is likely not the desired behavior. I've left comments in src/google/adk/agents/base_agent.py explaining the issue and suggesting a fix. I've also included a minor suggestion to improve code readability by combining nested if statements.
| if event := await self._handle_after_agent_callback(ctx): | ||
| yield event | ||
| # Skip after_agent_callback if the agent has handed off to another agent | ||
| if not ctx.agent_handoff_occurred: |
There was a problem hiding this comment.
There's a potential logic issue here. When agent1 hands off to agent2, ctx.agent_handoff_occurred is set to True. This correctly skips the after_agent_callback for agent1. However, agent2 inherits the context with this flag set to True, so its after_agent_callback will also be skipped, which is likely unintended. The callback should only be skipped for the agent that initiates the handoff.
A possible solution is to reset this flag at the beginning of each agent's run. You could add ctx.agent_handoff_occurred = False at the start of the run_async method in this file, after ctx is created. Since this change is outside the current diff, I cannot provide a direct code suggestion for it.
| if event := await self._handle_after_agent_callback(ctx): | ||
| yield event | ||
| # Skip after_agent_callback if the agent has handed off to another agent | ||
| if not ctx.agent_handoff_occurred: |
There was a problem hiding this comment.
Similar to my comment in run_async, there's a potential logic issue here that would cause the after_agent_callback to be skipped for the receiving agent as well as the transferring agent.
A possible solution is to reset this flag at the beginning of each agent's run. You could add ctx.agent_handoff_occurred = False at the start of the run_live method in this file, after ctx is created. Since this change is outside the current diff, I cannot provide a direct code suggestion for it.
| if not ctx.agent_handoff_occurred: | ||
| if event := await self._handle_after_agent_callback(ctx): | ||
| yield event |
There was a problem hiding this comment.
For better readability, you can combine the nested if statements into a single condition.
if not ctx.agent_handoff_occurred and (event := await self._handle_after_agent_callback(ctx)):
yield event| if not ctx.agent_handoff_occurred: | ||
| if event := await self._handle_after_agent_callback(ctx): | ||
| yield event |
There was a problem hiding this comment.
For better readability, you can combine the nested if statements into a single condition.
if not ctx.agent_handoff_occurred and (event := await self._handle_after_agent_callback(ctx)):
yield event2 participants
Please ensure you have read the contribution guide before creating a pull request.
Link to Issue or Description of Change
#3220
1. Link to an existing issue (if applicable):
2. Or, if no issue exists, describe the change:
If applicable, please follow the issue templates to provide as much detail as
possible.
Problem:
A clear and concise description of what the problem is.
Solution:
A clear and concise description of what you want to happen and why you choose
this solution.
Testing Plan
Please describe the tests that you ran to verify your changes. This is required
for all PRs that are not small documentation or typo fixes.
Unit Tests:
Please include a summary of passed
pytestresults.Manual End-to-End (E2E) Tests:
Please provide instructions on how to manually test your changes, including any
necessary setup or configuration. Please provide logs or screenshots to help
reviewers better understand the fix.
Checklist
Additional context
Add any other context or screenshots about the feature request here.