fix(acp): handle empty model response after function call

[codefromthecrypt]

Summary

Handle InvalidStreamError after sending functionResponse in ACP mode by treating it as a successful end of turn.

Details

Problem

In ACP mode, the Gemini model sometimes returns an empty response (thoughts only, no text) after a tool executes successfully. This triggers InvalidStreamError: Model stream ended with empty response text which propagates as an error to the ACP client, even though the tool completed successfully.

Root Cause

geminiChat.ts validates that responses have either tool calls or text content. When the model returns only thoughts after processing a functionResponse, this validation fails.

if (!hasToolCall) {
  if (!responseText) {
    throw new InvalidStreamError('Model stream ended with empty response text.');
  }
}

Solution

In zedIntegration.ts, catch InvalidStreamError after sending a functionResponse and treat it as a successful end_turn. The tool already completed - the empty text response is a model quirk, not a failure.

Related Issues

• ✕ [API Error: Model stream ended with an invalid chunk or missing finish reason.] #7851 - Original empty response bug report

How to Validate

1. Run Gemini in ACP mode
2. Trigger a file write operation
3. Approve the permission when prompted
4. Before fix: Error propagates to client after successful write
5. After fix: Clean end_turn, no error

Pre-Merge Checklist

• Updated docs/README (if needed) - N/A
• Added/updated tests (if needed) - Added unit test in zedIntegration.test.ts
• Noted breaking changes (if any) - None

[gemini-code-assist]

Summary of Changes

Hello @codefromthecrypt, 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 resolves an issue in ACP mode where the Gemini model occasionally sends an empty response following a successful function execution, which incorrectly triggers an InvalidStreamError. The changes introduce specific error handling in zedIntegration.ts to detect this scenario and interpret it as a successful completion of the turn, ensuring a smoother user experience and accurate state reporting after tool operations.

Highlights

• Error Handling for Empty Model Responses: Implemented a fix to gracefully handle InvalidStreamError when the Gemini model returns an empty response (thoughts only) after a successful function call in ACP mode.
• Treating Empty Responses as End of Turn: The system now treats an InvalidStreamError occurring after a functionResponse as a successful end_turn, preventing erroneous states for the client when the tool has already completed.
• New Unit Test Coverage: A new unit test has been added to zedIntegration.test.ts to specifically validate the correct handling of empty model responses after function calls.

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 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. ↩

[gemini-code-assist]

Code Review

This pull request addresses an issue where an empty model response after a successful tool execution was causing an error. The fix correctly handles this by catching the InvalidStreamError and treating it as a successful end of turn. The accompanying test case effectively validates this scenario. However, the error handling is a bit too broad and could mask other important errors. I've suggested a change to make it more specific.

[bdmorgan]

Hi @codefromthecrypt, thank you so much for your contribution to Gemini CLI! We really appreciate the time and effort you've put into this.

We're making some updates to our contribution process to improve how we track and review changes. Please take a moment to review our recent discussion post: Improving Our Contribution Process & Introducing New Guidelines.

Key Update: Starting January 26, 2026, the Gemini CLI project will require all pull requests to be associated with an existing issue. Any pull requests not linked to an issue by that date will be automatically closed.

Thank you for your understanding and for being a part of our community!

[codefromthecrypt]

Closing to focus on easier to repeat issues