Aren’t the generated tests superfluous since they just test whether the framework implements validation correctly? This should already be tested at the framework level and add no value?

This is an example of a current trend: tools that help both humans and AI. Since both perform cognitive work, anything that makes that work easier benefits both.

TypeScript-first API framework built on the idea that an API's specification, implementation, validation, and tests should never drift apart, because they are the same thing.

[dead]