This is duplicating the tests we already do in develop pointlessly. GitHub already detects merge conflicts that would change test results, and the rare case of the unclean merge does not seem close to happening and would be easily solvable by a follow-up PR anyways. (Or we could disable the build-test CI on develop.)