fix(core): remediate subagent memory leaks using AbortSignal in MessageBus

[abhipatel12]

Summary

Remediate subagent memory leaks caused by redundant MessageBus event listener accumulation.

Details

• Updated MessageBus.subscribe to accept an optional AbortSignal and implement manual listener cleanup upon abort events.
• Updated Scheduler to initialize an AbortController and trigger it during scheduler.dispose().
• Removed obsolete WeakSet deduplication logic from Scheduler to prevent bugs on reused buses.
• Added unit tests in message-bus.test.ts and scheduler.test.ts to cover the new behavior.
• Fixed unsafe types (Function and any) in the test files.

Related Issues

Related to #

How to Validate

Run the unit tests:

npm test -w @google/gemini-cli-core -- src/scheduler/scheduler.test.ts
npm test -w @google/gemini-cli-core -- src/confirmation-bus/message-bus.test.ts

Or run the full preflight:

npm run preflight

Pre-Merge Checklist

• Updated relevant documentation and README (if needed)
• 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
  • Linux
    • npm run
    • npx
    • Docker

[gemini-cli]

Hi @abhipatel12, 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!

[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 addresses potential memory leaks in subagents by implementing a more robust event listener management system. It integrates AbortSignal into the MessageBus subscription mechanism and the Scheduler's lifecycle, ensuring that event listeners are automatically cleaned up when they are no longer needed. This change enhances the stability and resource efficiency of the application by preventing the accumulation of redundant listeners.

Highlights

• MessageBus Listener Management: The MessageBus.subscribe method was updated to accept an optional AbortSignal, enabling automatic cleanup of event listeners when the signal is aborted. This prevents listener accumulation and potential memory leaks.
• Scheduler Resource Cleanup: The Scheduler now initializes an AbortController and triggers its abort() method during scheduler.dispose(). This AbortSignal is passed to MessageBus.subscribe calls made by the scheduler, ensuring that all associated listeners are automatically cleaned up when the scheduler is disposed.
• Obsolete Deduplication Logic Removed: The WeakSet deduplication logic (Scheduler.subscribedMessageBuses) in Scheduler was removed. This logic was intended to prevent duplicate listeners but could cause bugs when MessageBus instances were reused, and is now superseded by the AbortSignal mechanism.
• Improved Test Coverage: New unit tests were added to message-bus.test.ts and scheduler.test.ts to validate the new AbortSignal functionality and ensure proper listener cleanup.
• Type Safety Improvements: Unsafe types (Function and any) were fixed in the test files, improving code quality and maintainability.

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.

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. ↩

[github-actions]

Size Change: +1.19 kB (0%)

Total Size: 34 MB

Filename	Size	Change
./bundle/chunk-GN5PKLNC.js	0 B	-3.16 MB (removed)	🏆
./bundle/chunk-IIHVYCUQ.js	0 B	-3.47 kB (removed)	🏆
./bundle/chunk-SXO4DHDZ.js	0 B	-14.8 MB (removed)	🏆
./bundle/core-KCXUYDVW.js	0 B	-45.6 kB (removed)	🏆
./bundle/devtoolsService-D7JOOMDH.js	0 B	-28.4 kB (removed)	🏆
./bundle/gemini-AQWWYJFD.js	0 B	-552 kB (removed)	🏆
./bundle/interactiveCli-FANRV2CH.js	0 B	-1.64 MB (removed)	🏆
./bundle/oauth2-provider-REV7JLY6.js	0 B	-9.16 kB (removed)	🏆
./bundle/chunk-57VSEIPM.js	3.16 MB	+3.16 MB (new file)	🆕
./bundle/chunk-MTPSTLZ2.js	14.8 MB	+14.8 MB (new file)	🆕
./bundle/chunk-OUUI7CEF.js	3.47 kB	+3.47 kB (new file)	🆕
./bundle/core-O6KJVB4O.js	45.6 kB	+45.6 kB (new file)	🆕
./bundle/devtoolsService-L52P6JAG.js	28.4 kB	+28.4 kB (new file)	🆕
./bundle/gemini-QTORSUB4.js	552 kB	+552 kB (new file)	🆕
./bundle/interactiveCli-AV2XWW46.js	1.64 MB	+1.64 MB (new file)	🆕
./bundle/oauth2-provider-5DWPLIBC.js	9.16 kB	+9.16 kB (new file)	🆕

ℹ️ View Unchanged

Filename	Size	Change
./bundle/bundled/third_party/index.js	8 MB	0 B
./bundle/chunk-34MYV7JD.js	2.45 kB	0 B
./bundle/chunk-5AUYMPVF.js	858 B	0 B
./bundle/chunk-5PS3AYFU.js	1.18 kB	0 B
./bundle/chunk-664ZODQF.js	124 kB	0 B
./bundle/chunk-DAHVX5MI.js	206 kB	0 B
./bundle/chunk-ETSDHTGW.js	1.96 MB	0 B
./bundle/chunk-IUUIT4SU.js	56.5 kB	0 B
./bundle/chunk-RJTRUG2J.js	39.8 kB	0 B
./bundle/cleanup-CTXCUBW2.js	0 B	-856 B (removed)	🏆
./bundle/devtools-36NN55EP.js	696 kB	0 B
./bundle/dist-T73EYRDX.js	356 B	0 B
./bundle/events-XB7DADIJ.js	418 B	0 B
./bundle/gemini.js	4.97 kB	0 B
./bundle/getMachineId-bsd-TXG52NKR.js	1.55 kB	0 B
./bundle/getMachineId-darwin-7OE4DDZ6.js	1.55 kB	0 B
./bundle/getMachineId-linux-SHIFKOOX.js	1.34 kB	0 B
./bundle/getMachineId-unsupported-5U5DOEYY.js	1.06 kB	0 B
./bundle/getMachineId-win-6KLLGOI4.js	1.72 kB	0 B
./bundle/memoryDiscovery-YL35JRSL.js	980 B	0 B
./bundle/multipart-parser-KPBZEGQU.js	11.7 kB	0 B
./bundle/node_modules/@google/gemini-cli-devtools/dist/client/main.js	222 kB	0 B
./bundle/node_modules/@google/gemini-cli-devtools/dist/src/_client-assets.js	229 kB	0 B
./bundle/node_modules/@google/gemini-cli-devtools/dist/src/index.js	13.4 kB	0 B
./bundle/node_modules/@google/gemini-cli-devtools/dist/src/types.js	132 B	0 B
./bundle/sandbox-macos-permissive-open.sb	890 B	0 B
./bundle/sandbox-macos-permissive-proxied.sb	1.31 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
./bundle/sandbox-macos-strict-open.sb	4.82 kB	0 B
./bundle/sandbox-macos-strict-proxied.sb	5.02 kB	0 B
./bundle/src-QVCVGIUX.js	47 kB	0 B
./bundle/tree-sitter-7U6MW5PS.js	274 kB	0 B
./bundle/tree-sitter-bash-34ZGLXVX.js	1.84 MB	0 B
./bundle/cleanup-KI7MIIQV.js	856 B	+856 B (new file)	🆕

_{compressed-size-action}

[gemini-code-assist]

Code Review

This pull request refactors the Scheduler and MessageBus to utilize AbortSignal for managing listener lifecycles, replacing a WeakSet-based tracking mechanism. The MessageBus.subscribe method now accepts an AbortSignal to automatically remove listeners upon abortion, and the Scheduler uses an internal AbortController to manage its subscriptions, ensuring proper cleanup on dispose(). New tests have been added to validate this AbortSignal integration. A high-severity memory leak was identified in the MessageBus.subscribe implementation, where the abortHandler is not detached from the AbortSignal if unsubscribe is called before the signal is aborted. A detailed solution has been provided to address this by tracking abortHandlers and cleaning them up during unsubscribe.

[abhipatel12]

/gemini review

[gemini-code-assist]

Code Review

This pull request introduces AbortSignal support to the MessageBus.subscribe method, enabling automatic unsubscription. The Scheduler class was updated to use an instance-level AbortController for better resource management during disposal, replacing a static tracking mechanism. Feedback was provided regarding the use of a WeakMap in MessageBus, where using the listener as a key may cause cleanup functions to be overwritten if the same listener is reused across different event types or signals.

[abhipatel12]

/gemini review

[gemini-code-assist]

Code Review

This pull request introduces AbortSignal support to the MessageBus subscribe method, allowing for better lifecycle management of event listeners. It also updates the Scheduler to use this new functionality, ensuring that listeners are properly cleaned up when the scheduler is disposed. The reviewer identified a potential race condition in the subscribe implementation and provided a suggested fix to ensure the listener is correctly registered and cleaned up even if the signal is already aborted.

[jacob314]

nit: what does this test have to do with the new functionality?

[abhipatel12]

this was added b/c we didn't have any coverage to make sure that the listeners were actually working. I can remove if needed

[jacob314]

fine to leave. just checking

[jacob314]

Code Review

Reviewed by Jacob.

Minor edge case: The property tracks the cleanup functions using a . If the same listener function reference is subscribed to the same event multiple times (perhaps with different instances), the inner 's cleanup function will be overwritten by the last subscription.

While subscribing the exact same function to the same event multiple times is an anti-pattern, technically allows it. If this occurs, it would result in the previous s not being detached if is called manually before the signal aborts, potentially causing a minor memory leak of the 's event listener.

Consider tracking cleanup functions by something that uniquely identifies the subscription (or simply ignoring this if we enforce a single registration per listener function elsewhere).

[jacob314]

Code Review

Reviewed by Jacob.

packages/core/src/confirmation-bus/message-bus.ts

Minor edge case: The listenerToAbortCleanup property tracks the cleanup functions using a WeakMap<listener, Map<type, cleanupFunction>>. If the same listener function reference is subscribed to the same event type multiple times (perhaps with different AbortSignal instances), the inner Map's cleanup function will be overwritten by the last subscription.

While subscribing the exact same function to the same event multiple times is an anti-pattern, EventEmitter technically allows it. If this occurs, it would result in the previous abortHandlers not being detached if unsubscribe is called manually before the signal aborts, potentially causing a minor memory leak of the AbortSignal's event listener.

Consider tracking cleanup functions by something that uniquely identifies the subscription (or simply ignoring this if we enforce a single registration per listener function elsewhere).

[jacob314]