Break limits that only hold when requests arrive one at a time.
A check-then-act that is not atomic can be split by two requests that overlap. These labs are about
finding the one endpoint where that is true and proving it with real concurrency rather than a hunch.