It's the kind of code you should expect if you don't run a harness that includes review and refactoring stages.

It's by no means the best LLMs can do.