One interesting side effect of having a LLM write the thing including the README, is that the models tend to leave little hints of the authors intention or prompt as over-explained passages that make it obvious that AI help was used.

https://github.com/xsawyerx/curl-doom?tab=readme-ov-file#how...

eg. > A browser hitting the same URL gets a tiny landing page that just shows the one-liner

it’s subtle but once you notice it, it’s hard to miss.

As an aside, I feel like projects like this used to be really fun and impressive (I guess due to the fact that you’d think “Wow a human put their time into this wacky crazy thing”), whereas now you can have Claude consistently crap out something like this in 5 minutes, so it ruins the whole appeal to me…

Not to mention that the objectively bad practice of piping a curl call to bash is nowhere close to "playing doom via curl". It's almost as if they simply prompted "play doom with curl". In my experience, almost any overly-ambitious prompt ends similarly.