weekly usage burned in 1 day
Resolved 💬 3 comments Opened Apr 12, 2026 by tltrogl Closed Apr 12, 2026
💡 Likely answer: A maintainer (github-actions[bot], contributor)
responded on this thread — see the highlighted reply below.
What version of Codex CLI is running?
0.118.0
What subscription do you have?
plus
Which model were you using?
5.4 high
What platform is your computer?
_No response_
What terminal emulator and version are you using (if applicable)?
_No response_
What issue are you seeing?
It is telling me I burned through an entire weeks usage in 1 day.
What steps can reproduce the bug?
Uploaded thread: 019d7b33-0b06-7432-beaf-d9654a2706c3
What is the expected behavior?
_No response_
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
This usage burn issue highlights a critical blind spot in AI tool monitoring. We've observed that AI assistants can consume massive amounts of context and API calls on 'invisible work' - operations that appear productive but don't actually advance the user's goals.
One pattern we've documented is AI agents spending 60% of their quota on repeated failed attempts to parse malformed code or understand unclear requirements, while the user's actual task remains untouched. The assistant keeps trying different approaches, consuming more and more resources, without making progress.
What's dangerous about this is that it happens silently. The user sees 'activity' and assumes progress is being made, while the AI is burning through their usage limits on ineffective attempts. This creates a false sense of productivity that can lead to unexpected quota overages.
This is where Open Code Review complements traditional monitoring. While SonarQube and linters catch code quality issues, they can't detect that an AI has wasted 80% of your usage budget on failed attempts. We built usage efficiency tracking that correlates AI activity with actual user progress.
For anyone wanting to try: npx @opencodereview/cli scan . --sla L1
I have the same issue: