If it is small - it goes straight to PR If it is medium - open discussion, then PR If it is complex.... A MEP is required when X Experiments If everyone agrees -- then let’s just do it. If everyone doesn’t -- experiment How likely you will have to change it in future? How big will the impact be if you have to change it. Related: https://github.com/projectmesa/mesa/discussions/1909