Apart from what others have commented, IMO an “assertables” crate should not invent new predicates of its own, especially for domains (like math) that are orthogonal to assertability.