[Memory Leak] SSE streaming endpoint lacks disconnect detection and task timeout

[sanrishi]

Target

\�ackend/routers/ai.py\ : Lines 321–471 (\�nalyze_stream\ endpoint)

Vector

Resource Memory Leak (Performance)

Problem

The /ai/analyze_stream\ SSE endpoint creates an async generator that runs indefinitely without:

1. Client disconnect detection: No
   equest.is_disconnected()\ checks. Abandoned generators hold model references until completion.
2. Timeout protection: No max execution timeout. A frozen client keeps the generator alive, blocking resources.
3. Task cancellation: \�syncio.sleep()\ calls have no cancellation wrapping.

Changes

• Added _sse_disconnect_guard()\ helper checking
  equest.is_disconnected()\ before each yield
• Added _check_timeout()\ with 60s max window and early return
• Wrapped \�syncio.sleep()\ with \�syncio.wait_for()\ and 5s timeout
• Added try/finally cleanup logging on generator cancellation

Fixes #2868

[vercel]

Someone is attempting to deploy a commit to the ritesh Team on Vercel.

A member of the Team first needs to authorize it.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: c068ba8c-7743-4885-9411-49015f4f3dc1

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}