Please make codex explain it's train of thought, and what it's going to implement. What bug is it even tryna fix

Resolved 💬 3 comments Opened Jan 10, 2026 by skanda-vyas-srinivasan Closed Feb 23, 2026

💡 Likely answer: A maintainer (github-actions[bot], contributor) responded on this thread — see the highlighted reply below.

What feature would you like to see?

Codex keeps jumping to fix "problems" without even giving me a sliver of what it is even actually doing. Gets quite frustrating.

Additional information

_No response_

View original on GitHub ↗

3 Comments

github-actions[bot] contributor · 7 months ago

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

#9005
#6878

Powered by Codex Action

etraut-openai contributor · 7 months ago

Can you provide examples of prompts that you're giving Codex? You said that it's "jumping to fix the problem". If you'd like it to do some planning with you before jumping in to the problem, you can ask it to do so. Codex is designed to be highly autonomous, but it's able to be interactive too if you steer it in that directly.

If there's a particular session that you think epitomizes the feedback that you're giving, please use the /feedback command to upload the session data and post the thread ID here.

etraut-openai contributor · 6 months ago

The most recent model gpt-5.3-codex was trained to provide more details about its thought process.