CB-REV-0003: round 3, and three of four FATAL came from round 2's fixes

The pattern is now measured over three rounds: 5 fatal, then 3 (2 from the
previous round's corrections), then 4 (3 from them). The corrections are
not getting safer.

FATAL 1: round 2's short-cell assertion went into regulation.rs only.
attack-value.rs — which produced every number in CB-EV-0030's DARVO table
— still just warned, and the gate registered to close the finding claimed
the property for both.

FATAL 2, the sharpest of the three rounds: counting games proves they
STARTED. Stopping the engine after one round gives 200 games, all-zero
columns and exit 0 — byte for byte the signature CB-EV-0030 says the
instrumentation distinguishes from a real result. Both harnesses now
require every counted game to have reached an outcome over five rounds.

FATAL 3: round 2's `.csv` filter was applied to all three loops, so
catalog.yaml and rules_delta.yaml — whose missing digests were round 1's
finding — were recorded and then never compared, and never checked against
upstream at all. Only the parser loop filters now.

FATAL 4: five of six tiebreak comparators had no coverage. GR-E04's
tiebreak never executes in any scenario. All four are now covered and
mutation-verified; the Blame key needed compensating claims to be
reachable at all, since Blame also lowers the coalition score.

SERIOUS: "peak held" computed the same number as "peak assigned" for every
possible input — the real gap was that START_STRESS was an unchecked
constant, now read off the dealt state; cadence="none" was a pure
loophole, removed; sibling discovery swapped a hand-written list for
hand-written globs and missed metadata.json and VARIANT.md, both named in
the package's own changed_files — now walked, and it found them
immediately; and "~72,000 games" was unsourced, make panels runs 17,600.

Also separated two kinds of number that were presented alike: seats×games
is invariant, 363 and 29 vary 7.1%-11.5% across samples.

Round 4 owed. The conclusion is not that the work is nearly right — it is
that author-made corrections to measurement work should be assumed
defective until a fresh reader has attacked them.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

gates.toml

@@ -137,7 +137,7 @@ retire_if = "two passes run with no finding while artifacts keep growing — tha
 id = "CHAOS"
 name = "the chaos roll"
 target = ""
-cadence = "none"
+cadence = "manual"
 checks = "d8 on each tier declaration, 12-declaration calibration window (window 2, opened 2026-08-03; window 1 ran at d4)"
 added = "2026-07-30"
 review_by = "2026-11-30"
@@ -178,7 +178,7 @@ id = "CB-REV-0002/7"
 name = "variant panels"
 target = "panels"
 cadence = "all"
-checks = "the H1 measurement harnesses actually run, and a short cell fails rather than printing a number a reader must notice"
+checks = "both H1 measurement harnesses run; a short cell fails, and so does a cell whose games did not reach an outcome"
 notes = """
 Registered because the second adversarial review asked what the harness
 would report if the work silently stopped, and the answer was "green, and
@@ -186,4 +186,12 @@ nothing else": `attack-value` and `regulation` were wired into no target,
 so every figure in CB-EV-0030 and CB-EV-0031 came from a manual run of an
 ungated binary -- including the assertion added to catch short cells,
 which was unreachable from `make`.
+
+CB-REV-0003 #1 then found this `checks` line claimed a property that held
+for only one of the two harnesses: `attack-value`, which produced every
+number in CB-EV-0030's table, still only warned. And #2 found that
+counting games proves they STARTED: stopping the engine after one round
+gave 200 games, all-zero columns and exit 0 -- the signature CB-EV-0030
+says the instrumentation distinguishes from a result. Both now assert an
+outcome was reached.
 """