Error running remote compact task: {"detail":"Bad Request"}
Resolved 💬 3 comments Opened Jun 16, 2026 by rookepoole Closed Jun 29, 2026
💡 Likely answer: A maintainer (github-actions[bot], contributor)
responded on this thread — see the highlighted reply below.
What version of the Codex App are you using (From “About Codex” dialog)?
Version 26.609.71450
What subscription do you have?
I have Codex Pro 5x
What platform is your computer?
windows 11 version 26200.8655 64-bit operating system, x64-based processor
What issue are you seeing?
Every time I prompt now and 5.5 extra high goes to to automatically compact context its shows these errors {"detail":"Bad Request"}
Error running remote compact task: {"detail":"Bad Request"}
<img width="450" height="87" alt="Image" src="https://github.com/user-attachments/assets/98476939-6bd4-4a04-9191-7f78ccff3cdd" />
What steps can reproduce the bug?
I'm not sure what code snippet would reproduce it....
session ID: 019eaab4-61fe-72b3-b882-925d75deb5e4
Feedback ID: 019eaab4-61fe-72b3-b882-925d75deb5e4
What is the expected behavior?
I expect it to automatically compact the context
Additional information
_No response_
3 Comments
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
The expensive part of this failure is not just the 400, it is that the user loses the context of what compaction was trying to save.
A sturdier contract is to create a compact job receipt before the remote call, then return either a typed validation error with the rejected field or a resumable fallback path that preserves the pending summary locally. That way a bad remote request does not feel like the whole conversation state vanished behind one generic Bad Request.
Thanks for reporting this problem. Until recently, Codex used a separate endpoint and server-side logic for compaction. We recently switched to a more robust approach where the compaction logic is moved locally using the same endpoint as normal turns. This eliminates "remote compaction" errors and increases the stability and reliability of compaction operations. If you see further problems with compaction, please use
/feedbackand open a new bug report.