A consequence holds skeptically if it is true in every admitted model, in particular every stable model when using stable model semantics. Exhibiting one supporting stable model establishes only existence of support, not skeptical entailment.
New to topics? Read the docs here!