Board Pack
What changed since last time
Five movement states computed from two stored measurements, including the one that must never be reported as a fix.
Where a previous assessment exists, the pack states what moved. The comparison is computed from two stored measurements and is never inferred from the current assessment alone.
The five states
| State | Meaning |
|---|---|
| New | Failing now; the previous assessment never evaluated it |
| Reopened | Failing now; the previous assessment evaluated it as passing |
| Persistent | Failing in both. A changed resource set stays persistent, never closed |
| Closed | Failed before, evaluated now, and now passing |
| Not re-assessed | Failed before, and not evaluated now. Explicitly not closed |
Why the last row is the one that matters
A lost permission, a narrowed region scope or a scanner upgrade all remove controls from the denominator. Counting those as fixed would report a collection failure as progress.
A board told that fourteen findings closed, when in fact nine of them simply were not looked at, has been misled by arithmetic rather than by intent. The state exists so that cannot happen.
Comparability is gated
| Change since last time | Effect |
|---|---|
| The scoring constants changed | Composite and domain scores are withheld. Control-level movement continues, because control identifiers are stable across engine versions |
| The governance mapping version changed | Materiality band movement is withheld |
| A different account | The comparison is refused outright |
| The snapshot is missing, unreadable, or of an unknown version | The comparison is withheld, and the pack says which |
A first assessment is worded differently from a predecessor that could not be read. Those are different facts.
The comparison window
The stored snapshot is retained for 90 days by default. A re-assessment beyond that finds no predecessor and correctly renders as a first assessment. Nothing is fabricated. See Re-assessment.