Inconsistent use of tools, prompting user to run commands vs codex running them

Resolved 💬 2 comments Opened Oct 25, 2025 by ivangrynenko Closed Nov 11, 2025
💡 Likely answer: A maintainer (github-actions[bot], contributor) responded on this thread — see the highlighted reply below.

What version of Codex is running?

0.49

What subscription do you have?

Pro

Which model were you using?

gpt5-codex-low

What platform is your computer?

Darwin 25.0.0 arm64 arm

What issue are you seeing?

The front-end change in Drupal CMS requires compilation of front-end assets (JS, CSS, images, fonts, etc.). The model sometimes runs, but in most cases, it keeps telling me to run it, slowing down the process. Due to the time it takes for the model to complete a task, I cannot monitor the terminal and do something else, and due to this inconsistency, it takes extra time and effort to complete the task myself (run asset compilation and rebuild Drupal cache).

What steps can reproduce the bug?

Uploaded thread: 019a1a22-3371-72a3-9638-005495ab2d1a

Request to implement a style change and observe:

In some cases:

Next steps you may want to take:

  1. Run ahoy drush cr to ensure Drupal picks up the PHP change.
  2. Manually verify a waitlist-only event (e.g. /events/7-december-2025-sydney-3-days) to confirm the updated wording appears in the hero alert, sidebar, and both CTAs.

while in other cases:

Ran ahoy fe
  └ [fe] building  /app/docroot/themes/custom/sts
Ran ahoy drush cr
  └  [success] Cache rebuild complete.

What is the expected behavior?

Follow instructions and always execute the requested command.

Additional information

AGENTS.md contains detailed information on those commands

View original on GitHub ↗

2 Comments

github-actions[bot] contributor · 10 months ago

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

  • #5488
  • #5381

Powered by Codex Action

etraut-openai contributor · 9 months ago

Thanks for the bug report. We've shared the feedback with the team members who are responsible for training our codex models. We're tracking this feedback and similar reports so we can make the model more robust in future iterations.

If you see other model behavior that you'd like to report, please use the recently-added /feedback command in the CLI. That will provide us with additional details including logs and conversation text.