In Composer with any agent( including GPT-5.3), the UI shows “Taking longer than expected…” and the run appears to hang. The model stops mid-edit (e.g. code fades out on an incomplete line). It does not recover on its own. The only reliable way I’ve found to unblock it is to send a short follow-up such as “continue” in the input box. This has happened many times, not a one-off.
Steps to Reproduce
1\Open Composer in Agent mode with Composer 2 Fast.
2\Run a task that reads/edits files (e.g. JSP/JS as in the screenshot).
3\Wait until the status shows “Taking longer than expected…” below the streamed output.
Expected Behavior
The agent should either finish the step, show a clear error, or time out and offer a retry—without requiring a manual “continue” prompt.
Hey there,
This is a known issue that the team is actively working on.
A few workarounds that have helped other users:
Disable HTTP/2: Ctrl+Shift+P > Preferences: Open User Settings (JSON), add:
"cursor.general.disableHttp2": true
Then restart Cursor.
Reindex your project: Cursor Settings > Indexing & Docs > Codebase Indexing > Resync Index
Run Network Diagnostics: Cursor Settings > Network > Run Diagnostics — share the results here if the above steps don’t help.
If it happens again, please copy the Request ID right away (three dots menu in the top right of the chat > Copy Request ID) and share it here. That’s the key thing we need to investigate on the backend.
I’ve been surfing the forums for a while now trying to find a solution. I’ve tried them all. Nothing works.
This only happens on auto-mode, I know you say it’s a known issue and we should manually select a model, but since its out of Auto mode, this means I need to use On-Demand usage, which I dont want to do. Quite frankly, this is unacceptable. There is a degradation of service and we shouldn’t be the ones that pay for it.
You will have to excuse my tone. I am working weekends trying to rush development for clients and right now cursor seems like a real bottleneck. Don’t get me wrong, I love this IDE when it works, which right now it doesnt.
Saturday was a mess, could’nt use the auto mode starting from 14:00 EDT.
Sunday was fine.
Monday, fine until around 08:30 EDT. Then got the Taking longer than expected again.
I don’t have the stats that you guys have, but my guess would be that its related to too many people prompting or something. I guess fewer people work on sundays.
This issue is affecting almost all models; it doesn’t seem to occur in Plan mode, but it’s very frequent in Agent mode.
My team is considering canceling the Cursor subscription because this issue is causing a significant drop in productivity for most team members, which is extremely frustrating.
