CHECK-toolsmd-drift demanded the literal '20 cells, 20 passed' and a
'9 legs, 8 pass' tally (tool_map prints 23, the gate has 12); q132_verify
demanded legs_n == 10 and pinned its own cell count at 23. Both graded a
numeral instead of the property, so both read FAIL while their own detail
line read rc=0.
Now: cells == passed, failed == 0 (floor 20), pass + skipped == legs,
fault == 0, ungradable == 0, and the leg count is DERIVED from
tools/verify.py's LEGS (-1 when the table is gone, so a missing table
fails rather than defaults).
Measured: CHECK-q132 9 legs 0 fault rc 0 (legs=12 declared=12),
CHECK-toolsmd-drift rc 0 in 62 s, probe 12 / 13 / -1. Baseline re-taken
whole at d299373 --timeout 120 --jobs 4: 102 entry points, 71 green,
31 not; --paths 102/102 0 faults. Gate --fast: 12 legs, 11 pass, 0 fault,
gate OK.
The CHECK shipped with 6b7d750 ran the un-flagged repo gate: 107.93s wall,
82s of it the selftest-sweep leg. Inside the caller's harness it was killed
just after printing the section-4 header, so a green tree read as a failure.
Rebuilt as wizard-work/volt/CHECK-toolsmd-drift: per-leg timeout, the sweep
dropped from the budget (it is a wrapper over the cells graded directly here),
the leg tally asserted exactly, elapsed total printed. Non-vacuity measured by
three mutations on the real tree -- document drift (leg 1), uncommitted tool
drift (leg 2), index drift (leg 4) -- each rc 1, each restored clean.
volt