EDIT: Upon reflection, I do wonder if this was at least partially vibe coded, because LLMs don't have forethought and can't plan ahead for contingencies like that. Because it also seems like a no-brainer that it would lead to configuration failures if deployed on one of the sites that is blocked.
The dev was quite vocal about being against LLM usage. Most notably, I remember their April Fool's Day post about an update that adds a bunch of AI features. The post was obviously just a big joke, full of disparaging comments about LLMs. With that said, that was April (maybe even 2025, I can't remember). It's possible that their opinion had changed in the time since. But I doubt it, personally.
EDIT: It was 2025.
Canonical link: https://dubvee.org/post/3069664