i think if anyone starts using claude or any llm model regularly, the first thing they do when anything comes up is just open and start typing and instead of thinking
so eventually we get to know wether the model can do it or not,
But yeah these kinda directories help these tasks do in a better way