We are developing a small SDK that provided various helper methods to interact with our APIs, making life easier for daily tasks.
We want to provide some tests to make sure the wrappers are indeed talking to the API and getting results back, as well as handling any errors returned.
But how would we create isolated tests for these wrappers? Of course we can simply hook into a "sandbox" API that does return results, but I know that's not ideal.
To complicate matters, one API is a search endpoint, that will return different results for different queries; how would that even be mocked?