A test double used in test-driven development to simulate the behavior of a collaborator object, typically returning pre-programmed responses or values.
"The test stub provided canned responses to simulate interactions with an external service."
— @openai · February 25, 2024