I’ve created an open-source tool to do this. It’s called Tcases for OpenAPI. This automatically generates test cases that provide 100% “each choice” coverage for all API requests and responses defined in the spec. Typically, that also covers > 75% of the code lines in the API implementation.
I’m looking for feedback about how to make this tool better. In particular, I’d really like to hear suggestions about how to integrate Tcases for OpenAPI with the Prism mocking engine, to close the gap between the abstract test cases generated by the tool and the concrete input values needed to execute the tests.