Funnily enough, Web pages about the category don't use very accurate punctuation in their titles. Think of the category as "Punctuation: dash"... or, aptly, "Punctuation – dash".
Yes, it turns out to be a Unicode category just for dashes (well, actually dashes and hyphens). Unicode thrills!
Note that this isn't a strict match for an UUID, as you're including [g-z], and you aren't checking for the usual 8-4-4-4-12 format.
KM's regex engine does accept POSIX character classes, so you can use [:xdigit:] for 0-9a-f. No shorter, but perhaps more readable! So you could do something like:
I'm still a regex noob -- most of what I've learnt has been from trying new things in this Forum, and seeing how others solve the same problem. The big thing, IMO, is practice at spotting patterns in the first place -- how to match/use them is then just a google away
This regex can certainly be improved. For example, we can make the "inside" group non-capturing to make it slightly more efficient and -- bonus! -- suppress the unnecessary KM variable field:
\b([[:xdigit:]]{8}(?:-[[:xdigit:]]{4}){3}-[[:xdigit:]]{12})\b
^^
make group non-capturing
I'm sure others here can improve it still further.