Every finding SiteCMD shows you has a severity, a confidence, and a status. Together they decide where the issue lives in the list, whether it caps your score, and what action makes sense next.

Severity

Severity is how bad the thing is if it’s real. Four levels:

Severity	What it means
Critical	A confirmed critical caps your score at 49. Usually a security hole, a broken page, an exposed secret, or something that would block launch. Fix or mark not applicable.
High	A confirmed high caps your score at 79. Functional gaps, missing security headers, slow load on a key path, broken canonical tags, things that hurt users right now.
Medium	Things that should be fixed but aren’t blocking. Most accessibility issues, polish problems, minor SEO gaps.
Low	Nice-to-haves and weak signals. They count toward the score, but only a little.

Severity is set by the check itself based on what it found. It’s not a guess at how much effort the fix will take.

Confidence

Confidence is how sure SiteCMD is that the finding is real. Three levels:

Confidence	What it means
Confirmed	SiteCMD observed direct evidence. The HTML literally has the problem, the lockfile literally has the bad version. No ambiguity. Counts at full weight, can trigger score caps.
High	Strong automated evidence. The signal is reliable in most cases but there’s a small chance of a false positive. Default for findings without an explicit confidence. Counts at 0.85× weight, can trigger caps.
Needs review	A possible issue that should be reviewed before it blocks a launch. The signal is real but might not apply to your situation. Counts at 0.55× weight. Does not trigger score caps.

If a critical issue is flagged at needs review confidence, your score doesn’t get capped at 49. SiteCMD won’t pull the alarm on something it isn’t sure about. You can promote it manually (or fix it, or mark it not applicable) once you’ve looked at it.

Status

A finding doesn’t sit in one state forever. Every issue has a status that controls whether it counts toward your active list and your score.

Status	What it means	Counts toward score?
New	Just discovered, or recently reopened. Shows in the active list.	Yes
Snoozed	Paused until a date you picked. Drops out of the active list until the date passes.	No, until snooze ends
Ignored	Permanently dismissed as not applicable. Hidden from the active list.	No
Blocked	Flagged with a reason. Used for “waiting on external dependency” or “design decision pending”.	No
Verified	Marked fixed. Removed from the active list. If a later scan finds the same issue, it gets reopened.	No

The only status that counts toward your score is New. Everything else (snoozed, ignored, blocked, verified) is excluded from active counts and the score calculation. This is intentional: the score is meant to reflect the work that’s still on your plate, not work you’ve explicitly set aside, decided isn’t your problem, or already finished.

A “verified” issue is removed from your active list, but SiteCMD doesn’t take your word for it forever. The next time a scan runs the relevant check, the issue either stays verified (the check passes) or pops back to “new” (the check fails again). Self-reported fixes that don’t actually work get caught.

A “blocked” issue stays out of the score while it’s blocked, but the Blocked filter on the Issues page shows you everything in that state at a glance, with the reason you set. It’s a tracked-but-paused bucket: useful when something is genuinely out of your hands and you don’t want it polluting your active list, but you also don’t want to forget about it.

How findings are ranked

The Issues list sorts by impact, not severity alone. Impact is the per-issue penalty: the same number that goes into the SiteCMD Score. This means:

A confirmed critical at high impact sits above an unconfirmed critical.
A high-confidence high (12 points) sits above a confirmed medium (5 points).
An issue that shows up on 5 pages sits above the same severity issue that shows up on 1.

If you start fixing from the top of the list, you’re fixing in the order that will most move your score. For working through findings instead of just picking the worst, see Triaging issues.