There are positive, negative and edge cases.

They are generated by AI, so unit tests will be implemented by case definitions.