[foltik]

Based on the recent leaks, their system prompt explicitly nudges the model not to do anything outside of what was asked. That could very well explain why it’s not fixing preexisting broken tests.

“Don't add features, refactor code, or make "improvements" beyond what was asked.”

https://www.dbreunig.com/2026/04/04/how-claude-code-builds-a...

[hakanderyal]

And it's very valid. Because otherwise you would ask Claude to trim a tree and it would go raze the whole forest and plant new seeds. This was the primary pain point last year, especially with Sonnet.

[cmrdporcupine]

Whatever prompting OpenAI has with Codex / GPT 5.4 seems superior here then.

It's very surgical and careful around incremental refactoring, etc. but it also doesn't avoid responsibility.