I am working on an AI agent that writes unit tests. Soon we will be launching officially. How it works is it contextualizes the methods and dependencies in a codebase and writes unit tests for intended functionalities/business logic.
The way we integrate ourselves into the development lifecycle is by being called on several platforms (slack GitHub and in IDE).
I was wondering what are your critiques of this workflow? To give a concrete example before you commit, you mark the methods you need tests for or in IDE you just push generate tests button and we generate it.
I am only curious about your take on this UX.