fix(core): set temperature to 1 on retry in sendMessageStream

[SandyTao520]

TLDR

Sets the temperature to 1 in GenerateContentConfig during retries within sendMessageStream.

Dive Deeper

When sendMessageStream encounters invalid content and triggers a retry, increasing the temperature to 1 encourages the model to generate different output, potentially resolving the issue that caused the initial failure.

Reviewer Test Plan

Reviewers can validate this by mocking generateContentStream to fail on the first attempt with invalid content and succeed on the second, verifying that the config passed to the second call has temperature: 1.

Testing Matrix

	🍏	🪟	🐧
npm run	✅	❓	❓
npx	✅	❓	❓
Docker	❓	❓	❓
Podman	❓	-	-
Seatbelt	❓	-	-

Linked issues / bugs

[gemini-code-assist]

Summary of Changes

Hello @SandyTao520, 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 introduces a crucial enhancement to the sendMessageStream functionality within the core package. It addresses scenarios where generative model calls might fail due to invalid content by implementing a retry strategy that modifies the temperature parameter. By increasing the temperature to 1 on subsequent attempts, the system aims to prompt the model to generate more varied and potentially valid responses, thereby improving the robustness of content generation. The change is accompanied by a dedicated test case to validate its effectiveness.

Highlights

• Retry Mechanism Enhancement: The sendMessageStream function now explicitly sets the temperature parameter to 1 in the GenerateContentConfig when a retry attempt is initiated due to invalid content. This encourages the generative model to produce a different output, aiming to resolve the initial failure.
• New Test Case: A new test has been added to geminiChat.test.ts to verify that the temperature is correctly set to 1 during retry attempts, ensuring the intended behavior.

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 correctly implements a retry strategy for sendMessageStream by increasing the model's temperature to 1 on subsequent attempts after an invalid content failure. This should help in getting a different, hopefully valid, response from the model. The implementation is sound and includes a good unit test to verify the behavior. I have one suggestion to improve the robustness of the code against potentially invalid input types for the configuration object, which could prevent runtime errors for JavaScript users.

[gemini-code-assist]

The spread operator ...currentParams.config can throw a runtime TypeError if currentParams.config is not an object (e.g., if it's a string or any other primitive). While TypeScript provides some safety, JavaScript users could pass an invalid type for the config property, leading to an uncaught exception. To make this more robust, you should ensure currentParams.config is a spreadable object before using the spread operator.

Suggested change

	currentParams.config = {
	...currentParams.config,
	temperature: 1,
	};
	currentParams.config = {
	...(currentParams.config && typeof currentParams.config === 'object' ? currentParams.config : {}),
	temperature: 1,
	};

[github-actions]

Size Change: +255 B (0%)

Total Size: 17.8 MB

ℹ️ View Unchanged

Filename	Size	Change
./bundle/gemini.js	17.7 MB	+255 B (0%)
./bundle/sandbox-macos-permissive-closed.sb	1.03 kB	0 B
./bundle/sandbox-macos-permissive-open.sb	830 B	0 B
./bundle/sandbox-macos-permissive-proxied.sb	1.31 kB	0 B
./bundle/sandbox-macos-restrictive-closed.sb	3.29 kB	0 B
./bundle/sandbox-macos-restrictive-open.sb	3.36 kB	0 B
./bundle/sandbox-macos-restrictive-proxied.sb	3.56 kB	0 B

_{compressed-size-action}

[mrcabbage972]

/patch release

[github-actions]

✅ Patch workflow(s) dispatched successfully!

📋 Details:

• Channels: stable,preview
• Commit: 8dc397c0a50fed77d072bf889140f5f8876bcb30
• Workflows Created: 2

🔗 Track Progress:

• View patch workflows
• This workflow run

[github-actions]

🚀 Patch PR Created!

📋 Patch Details:

• Channel: stable → will publish to npm tag latest
• Commit: 8dc397c0a50fed77d072bf889140f5f8876bcb30
• Hotfix Branch: hotfix/v0.9.0/0.9.1/stable/cherry-pick-8dc397c/pr-10866
• Hotfix PR: #11367

📝 Next Steps:

1. Review and approve the hotfix PR: #11367
2. Once merged, the patch release will automatically trigger
3. You'll receive updates here when the release completes

🔗 Track Progress:

• View hotfix PR #11367

[github-actions]

❌ Patch creation failed!

There was an error creating the patch release.

🔍 Troubleshooting:

• Check the workflow logs for detailed error information
• Verify the commit SHA is valid and accessible
• Ensure you have permissions to create branches and PRs

🔗 Links:

• View workflow run

[github-actions]

🚀 Patch Release Started!

📋 Release Details:

• Channel: stable → publishing to npm tag latest
• Version: v0.9.0
• Hotfix PR: Merged ✅
• Release Branch: release/v0.9.0-pr-10866

⏳ Status: The patch release is now running. You'll receive another update when it completes.

🔗 Track Progress:

• View release workflow

[skeshive]

/patch preview

[github-actions]

✅ Patch workflow(s) dispatched successfully!

📋 Details:

• Channels: preview
• Commit: 8dc397c0a50fed77d072bf889140f5f8876bcb30
• Workflows Created: 1

🔗 Track Progress:

• View patch workflows
• This workflow run

[github-actions]

❌ Patch creation failed!

There was an error creating the patch release.

🔍 Troubleshooting:

• Check the workflow logs for detailed error information
• Verify the commit SHA is valid and accessible
• Ensure you have permissions to create branches and PRs

🔗 Links:

• View workflow run

[github-actions]

❌ Patch creation failed!

There was an error creating the patch release.

🔍 Troubleshooting:

• Check the workflow logs for detailed error information
• Verify the commit SHA is valid and accessible
• Ensure you have permissions to create branches and PRs

🔗 Links:

• View workflow run