P-M2-2: the false-positive scanner pays for itself early

id:        P-M2-2
filed:     2026-07-17, from spec 2105 doc 10 section 6.4, before M2 slice two starts
suite:     the recorded passing traces the scanner replays
tier:      whatever tiers the recorded traces span
engine:    the tomo-oi guard family, scanned offline
band:      floor: at least one latent guard over-trigger found during slice two's
           development, before the scanner even reaches CI.
           ceiling: the over-triggers found are in guards written during experiment
           rushes against failing traces.
mechanism: guards written in a rush against failing traces have never been
           systematically run against passing ones, so the first systematic pass over
           the passing corpus should surface at least one guard that fires where it
           should not.
settled by: the first full scanner run over the recorded passing corpus, during M2
           slice two.
result:    (open)
verdict:   (open)

A clean first scan would itself be a mild surprise worth a verdict: it would say the guard family's trigger conditions were written more carefully than their development history suggests.