I agree with the sentiment but LEO constellations like Starlink can and have been disrupted, via sub-orbital jamming. Not to mention that in actual large conflict surface to LEO missiles will simply destroy large amounts of satellite constellations.
And as the other user mentioned, no country at the moment has the kind of stockpile of ASAT weapons needed to wipe the constellation (plus, due to orbital dynamics, there's a limit to how quickly they can take out satellites).
Between trying to wipe the constellation and jamming it, it'd be far more cost effective to jam even accounting for the higher power requirements/lower jamming range.
There would also be other interesting options like capturing and using enough terminals to force the entire cell to be disabled. That has been one of the challenges SpaceX has had to deal with near the frontlines in Ukraine.
Even modest investment in better tracking could massively improve crash avoidance. It would take far more then a handful of sats to truly impact the functioning of the network. And even more to complete take it out.
Russia is working on everything if you believe their marketing. I seriously question if any work on 'orbital nuclear weapons' is anything other then marketing. And its questionable how effective that would actually be.
This isn't as easy as people think. A country like Russia might have some readiness of nuclear weapons. And maybe a small readiness of anti-sat weapons, but not anywhere close to enough to attack a network like Starlink. Preparing for something like that simply wasn't a thing anybody considered necessary until 2020 and Russia certainly hasn't invested huge amounts of money in that since then and given their recent success with rocket development, I not sure how effective it would have been even if they had.