# COM-014 Lab

Run `python run_checks.py`. Fixtures are synthetic and offline. Rebuild each bounded feedback decision from metric semantics, instrumentation, collection quality, window/segment, baseline and the predeclared decision rule.