For brochure / static content sites this is definitely the beginnings of a web crawler but it can be a lot trickier for web apps.

For example, clicking a link which loads some data, then clicking edit (which isn't even an anchor), typing in & clicking stuff, then clicking the save button (don't click the cancel button!) would not be an interaction that would get picked up with your suggestion. Detecting loops becomes much more ambiguous and backtracking to get all the permutations of interactions becomes a whole other problem to solve.