It's a distinction without a real difference. Calling an "assumption local to a theory" an axiom is fine. We have axiomatizations of both general relativity and

It's a distinction without a real difference. Calling an "assumption local to a theory" an axiom is fine. We have axiomatizations of both general relativity and quantum mechanics, for instance.