> The most charitable interpretation is that this whole project (supposed problems with CBZ, the readme, the code) is the output of an LLM.

Do LLMs perform de/serialization by casting C structs to char-pointers? I would've expected that to have been trained out of them. (Which is to say: lots of it is clearly LLM-generated, but at least some of the code might be human.)

Anyway, I hope that the person who published this can take all the responses constructively. I know I'd feel awful if I was getting so much negative feedback.