Inwom is a security tool designed to detect coordinated behavior among AI agents and machines by analyzing collective activity patterns rather than individual events. It reconstructs how agents organize, share capabilities, and act as coordinated systems using observable evidence from trusted infrastructure, distinguishing legitimate automation from potential attacks.