Model went on an infinite loop when running /goal

Resolved 💬 4 comments Opened May 11, 2026 by Microck Closed May 11, 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?

codex-cli 0.130.0

What subscription do you have?

Plus/Business

Which model were you using?

gpt-5.5

What platform is your computer?

Linux 6.8.0-1050-oracle aarch64 aarch64

What terminal emulator and version are you using (if applicable)?

tmux

What issue are you seeing?

Codex CLI /goal got stuck in a repeated “blocked” state: it kept saying live verification was required but unavailable/unwaived, while also refusing to call update_goal or take a next action. I had to interrupt the loop manually.

What steps can reproduce the bug?

thread ID 019e1692-3e53-7243-9831-1f5847a4cc23

What is the expected behavior?

_No response_

Additional information

<img width="429" height="679" alt="Image" src="https://github.com/user-attachments/assets/d256e706-1ffc-4cb6-802d-07434e079b26" />

View original on GitHub ↗

4 Comments

github-actions[bot] contributor · 3 months ago

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

  • #21786
  • #22090
  • #22132

Powered by Codex Action

etraut-openai contributor · 3 months ago

Yes, this is the intended behavior when /goal is enabled. The /goal feature pushes the model to keep trying until the goal is achieved.

Tuo-Luo · 3 months ago

I documented a related workaround here: #23067.

The approach is to make repeated blocking conditions part of the goal completion criteria up front, so /goal can terminate the continuation loop when it keeps hitting the same blocker, without treating the blocked action itself as completed.

bobozai2019 · 1 month ago

I meet the same problem , does it has been fixed?