Four result patterns, not one.

Pattern A

Option comparison

Every candidate stays visible on the result table — including the failures, with the reason each one failed. The selected row is marked, but nothing is hidden to make the table look cleaner.

Pattern B

Clause checklist

Requirement set against design, with a status per clause. There's no "selected" option here — the tool is checking compliance, not picking a winner.

Pattern C

Judgement record

A table with a justification column, outstanding actions, and a signature block. The output is a document, not just a number.

Pattern D

Direct calculation

One input set, one computed answer, with the formula and the standard it's drawn from shown alongside it. There's no candidate to select and no clause to check — just a number you can trace back to its working.

Seed vs verified

Where a figure rests on a seed model rather than a licensed table set, the result says so — a CHECK chip, not a silent PASS. A tool admitting uncertainty is doing its job, not failing at it.