To expect manual or automatic tests to cover all changes is naive. If people start to make changes that is not related to the actually task. The risk of missing

To expect manual or automatic tests to cover all changes is naive. If people start to make changes that is not related to the actually task. The risk of missing something is kinda obvious. You (and the op) is barking up the wrong tree. Surely you have to understand it is not the fault of git that ends you up in this stupid situation. It is a broken process. Edit: grammar, not all :)