Hello everyone, I’ve been testing the Github Classroom autograder for use in several courses that I’m conducting this semester. One thing that alarmed me was that, after setting an assignment (with test cases) and inviting test accounts, I could modify the json parameters for the test script from the student accounts. These changes were reflected in the automated evaluations as I had successfully changed the total score of an assignment from 10 to 100. How may I prevent this?
Moreover, is there a way to prevent students from modifying boilerplate (starter) code or automatically rejecting such commits? Can I somehow run the automated tests in such a way that the unit test code is in a repository separate from all the student repositories?