A skill should only document behaviour the LLM didn't/couldn't exhibit on its own.

So you take your failed case (eg. working with gdb or whatever), write a skill and then test for that failed case.

There are also skills that help LLM do the thing it can do without the skill, but faster (by cutting out unnecessary discovery). I guess for such skills the fail case is "being slow"?

I imagine many fail cases can burn a lot of tokens/usage/time because failing LLMs can be very persistent. Maybe some upper bound (turn count, timeout) would help too.