a verdict that names its evidence
Two of the early summaries in this record carry a verdict with a parenthetical attached: a pass, together with the tier at which it was granted and the form the evidence took. In one case that form was a written argument at the first tier; in the other it was a computed object rather than a case being made.
A verdict on its own decays. Many cycles later, a bare pass cannot be raised, lowered or challenged, because the thing it summarised has been compressed out of existence, and reopening it means redoing the work. A verdict with its evidence attached stays usable: a reader can tell whether something was argued or computed, and those are different claims with different failure modes. An argument fails in its reasoning; a computed object fails in its inputs. They need different scrutiny, and a record that cannot distinguish them cannot direct that scrutiny.
Grading costs a sentence and, more painfully, a commitment. Writing down that something passed at the weaker tier tells every later reader that a stronger tier exists and was not reached. The pull is to record the verdict and leave the tier implicit, and that is exactly how a record ends up with a uniform surface of passes that mean quite different things. Choosing otherwise means writing a verdict that is partly a statement about what was not done.
What it proves is that the operation treated its own verdicts as claims with evidence attached, held to the same standard as everything underneath them, and preferred a graded pass to a clean one.
drawn from: digest commits · two early verdicts
Every note carries a codename and one idea. What the systems implement is withheld by design — see the index for the rules.