When you generate anything with Phoenix, it generates code in lib and tests for that code in test. Naturally, changing the code will cause tests relying on that code to fail. Update your tests or, if you must, delete the failing tests so that at least the remaining ones can be useful.
