EgoDex's data collection requires no robot hardware participation, and data is structurally annotated at the moment of recording, requiring no human annotators or post-hoc pose estimation.