Bug: /goal floods session with repeated (remote compact) usage-limit errors

Resolved 💬 2 comments Opened May 17, 2026 by muzahidulislamhadi Closed May 17, 2026
💡 Likely answer: A maintainer (github-actions[bot], contributor) responded on this thread — see the highlighted reply below.

Summary

When running /goal in Codex, if usage limits are reached during the run, the session thread is flooded with repeated identical remote compact task errors instead of showing a single deduplicated/terminal usage-limit message.

Impact

This makes the session noisy and hard to inspect, obscures the actual run state, and can produce dozens of duplicate messages after the actionable error has already been reported. This should be handled immediately because it degrades the Codex run UX exactly when users need a clear failure state.

Observed behavior

The run thread repeatedly prints the same message:

■ Error running remote compact task: You've hit your usage limit. Upgrade to Pro (https://chatgpt.com/explore/pro),
visit https://chatgpt.com/codex/settings/usage to purchase more credits or try again at 10:29 PM.

This message was emitted many times in the same session.

Expected behavior

After the first usage-limit failure, Codex should stop retrying or suppress duplicates, then show one clear terminal/status message, for example:

Usage limit reached. Remote compact task stopped. Try again after 10:29 PM or purchase more credits.

Suggested fix

  • Treat usage-limit errors from remote compact as terminal for that task/session path.
  • Deduplicate identical errors in the session log.
  • Add retry/backoff guards so quota failures cannot flood the thread.
  • Optionally collapse repeated internal errors behind a single status entry.

Severity

High / requires immediate fix: repeated quota errors flood active run threads and obscure the user-facing failure state.

View original on GitHub ↗

2 Comments

github-actions[bot] contributor · 3 months ago

Potential duplicates detected. Please review them and close your issue if it is a duplicate.

  • #23159

Powered by Codex Action

etraut-openai contributor · 3 months ago

Thanks for the bug report. I'm working on a fix for this.