Before leaving the site, an engineer runs read-only show commands covering the platform, interfaces, enclosures, and capacity, and archives the output with the deployment record. Why is that baseline the most valuable thing captured during first access?

A read-only baseline is the control group for every future comparison. The two-a.m. questions — should this interface be down, was the enclosure count always this, has usable capacity been shrinking for weeks — are answerable only when a snapshot of exactly those views exists from the day the system was certified working; the commands change nothing, so taking them at commissioning carries no risk and buys all of the diagnostic leverage. Replaying saved show output to rebuild a replacement fails by concept: those views are human-readable reports, not configuration syntax, and a rebuilt system needs applied configuration plus restore media, not archived text. Audit exemption does not exist — a baseline may support compliance evidence, but the device is scanned on the same schedule as everything else. Suppressing future alerts inverts the purpose of monitoring entirely; the baseline sharpens alert interpretation, it does not mute it. The exam trap here is treating baseline collection as optional paperwork rather than part of the access task itself: this is the last moment the state is known-good and the only engineer who saw it is the one holding the terminal. Operational check: capture platform, interface, enclosure, and capacity views, store them with the deployment record, and reference them in the runbook so the first responder after an incident opens the baseline before the speculation.