[Bug] Opus model code quality degradation on iterative fix cycles

Question

**Bug Description**
Feedback: Opus-4.7 code quality degradation on simple scripts
  
  Codex (GPT-5.5) keeps reviewing and finding errors in Opus-4.7's code. The review findings
  are passed back to Opus-4.7 for fixing, but after each fix, new errors appear. This is a
  simple shell script — 193 lines, 668 words, nothing complicated.

Why is the quality degraded on straightforward tasks like this?

**Environment Info**
- Platform: linux
- Terminal: vte-based
- Version: 2.1.150
- Feedback ID: 0