Opus 5.5 has one very serious, catastrophic problem that nobody is talking about: it's unfit for code review.
It's been made so "safe" and so restricted that, out of a hundred confirmed critical issues, it will flag two at most. Opus 5.5 writes completely insecure software that can't be shipped to production
I'm convinced Anthropic has gone overboard with restrictions and safety. This needs to be fixed urgently
Keep this in mind!
1
4 comments
@Gegam245074 Use Astra to double check everything, its been working well so far finds so many issues all the time!
@CutfeedTV Yes, Astra and the other GPT models really are the best for reviews
@Gegam245074 I have found and fixed hundreds of bugs in an open source library with Opus 5 and 5.5 -- I never observed what you are talking about. Provide a little more detail about exactly what you are doing -- maybe I can provide some tips.
Opus 5.5 is a great model and writes the best code of them all. But when it comes to bug hunting, it's seriously limited. Run Astra for a review and run Opus 5.5 on the same project. Astra will give you 30 critical bugs, while Opus finds 5. Then ask Opus to check Astra's output, and Opus will confirm every single finding