Daily checks should cover not only your own key pages but also the status pages and notices of dependencies: a degradation at the CDN, payment gateway, logistics API or analytics script shows up as 'our pages are slower' or 'some fields are missing'. Keep a list of these dependencies with each one's status entry and owner.
When a dependency looks abnormal, first separate 'their incident' from 'our config drift': re-scan a fixed sample page that does not use that service. If the sample is fine, the fault is likely on their side; if both are abnormal, check our own change log first.
This article covers inspection scope and attribution order only; it makes no quality judgement about any dependency. Record status as it is — when a dependency is fine, log 'normal' too, instead of leaving a trace only when things break.