CrUX data is not based on a crawler but on real Chrome users who agree to share anonymised data with Chrome.
That's what makes it so powerful—these are what your actual users are experiencing (or at least a subset of them using Chrome and meeting the other criteria).
It's also what makes it a bit more complicated, in that the users and their conditions can change (e.g. one week you make get a lot more of X type users, another week less) so your real-world Core Web Vitals numbers can change even if your website doesn't. But that's still what your real users over that period experienced.