What it is
A verbose health verdict that surfaces only when something needs attention. It aggregates deploy status, uptime, SSL/DNS, and placeholder-serving detection into one banner that says — in plain English — what's wrong, what the public is seeing right now, and how to fix it. Failed builds are classified (SSR config missing, dependency install failed, prerender timeout…) instead of dumping raw logs. When a site goes down or recovers, cite-met emails the owner and team with the same plain-language verdict — at most once per incident, so no alert storms.
Site Health panel explaining in plain language what needs attention (demo data, illustrative)
Plain-English verdicts
What's wrong, what the public sees now, and how to fix it.
Classified build failures
SSR config, dependency, prerender — not raw log dumps.
At-most-once alarms
Outage and recovery emails without alert storms.
Why it matters
Non-engineers understand it and engineers skip straight to the cause. And proactive outage comms turn a reliability event into a trust-building moment — your customers hear it from you first, not from theirs.