One of the things LLM shines. For double checking the command explanations, I ask commands to grep the sections from manual instead of relying LLM output blindly.