Crisper can snip genomes in particular spots, we probably have the genome of corona virus. Viruses probably don't work if their genome is chopped up. What would happen if you loaded random chunks of the coronavirus genome into crispr, and injected it into someone with coronavirus? I don't know much biology, this may be completely stupid, but if there is someone out their with enough biology knowledge and biohacking skills...
Virus DNA is wrapped in protien shells in the intercellular fluid. You need some way to get your CAS9 into cells, like wrapping it in a virus protien shell, to hijack the viral delivery mechanism. We can use coronavirus itself. Get cas9 targeted to the coronavirus genome, put it within a coronavirus protien shell, and inhale the spray. (You don't need the defences in all your cells, just in your lungs.) The virus injects cas9 in to your cells, if you don't get coronavirus, the cas9 does nothing. If you get it in a cell, the cas9 chops it up. The virus protien shells help train your immune system.
Mutation rate, most of the time, one snip will kill it, sometimes it could mutate. Don't set the crispr to a part you don't want to mutate. Set it somewhere where mutations are harmful in the coronavirus genome, do these viruses have a part that looks the same?