Our longterm plan is to classify knot modules and pairings by utilising the power of computational number theory. The first step in this is to define invariants for which any given value arises from only finitely many modules: this is the purpose of the present paper.