refinement-type-checker

Refinement types verify code properties by attaching logical predicates to standard types. This Skill checks pre- and post-conditions by passing your predicates to the Z3 SMT solver for automated subtyping evaluation.

Refinement types offer a lighter alternative to full dependent types for contract checking. They provide a dependent type-lite subset that enforces mathematical properties via SMT solvers rather than requiring a fully dependent type system.

You enforce function pre- and post-conditions by defining logical predicates for your types. The Skill utilizes the Z3 SMT solver to evaluate these predicates and automatically verify that your code adheres to the specified contracts.

Yes, you need both Python and the Z3 solver to use this refinement type checker. Python runs the verification scripts, while Z3 acts as the underlying SMT solver required for predicate evaluation and subtyping checks.

Yes, property verification can guarantee a function returns a non-negative integer. You refine the return type with a logical predicate like greater than zero, and the Z3 solver proves the function output always satisfies this condition.

Dependent type-lite features limit program analysis to properties expressible as logical predicates solvable by Z3. It is a lighter alternative to full dependent types, meaning highly complex or undecidable mathematical constraints may not verify successfully.