phase9: ledger — cycle-4 363 bodies, rare-epilogue model confirmed, bad-claim isolation

phase-ends/CURRENT_PHASE.md

@@ -510,3 +510,24 @@ that ALSO contain a move pseudo — those block on the mutual exclusion.
 LESSON: the coordinator over-generalized a class consequence from a census;
 worker B's empirical falsification is the correct checks-and-balances that
 makes the team structure work.
+
+## Cycle 4 (continued) — 363 bodies; rare-epilogue model confirmed
+
+| Cycle | New bodies | Total bodies | New regions | Total regions | Result |
+|---|---|---|---|---|---|
+| c4 (to date) | +64 | **363** | +64 | **372** | green merges |
+
+- Worker B: 26 new this cycle. The rare-epilogue three-way model got its
+  textbook confirmation: 0x800F8928 (sub-case b) behaves exactly as
+  predicted (maspsx off fixes the 4-byte epilogue, every move becomes or;
+  mutual exclusion). Three independent confirmations now: 0x80021FF4 (a),
+  0x800F8928 (b), 0x800FBDC0 (c). Pass-through 5th flavour
+  (discard-an-argument-outright). Stale-request pruning automated by B.
+- Worker C: 4 merged; 0x8002E3FC DROPPED from the merge (claimed but
+  LENGTH-MISMATCH 88 vs 80 against the tracked registry — reported back
+  with evidence).
+- GATE INCIDENT methodology held: a single bad row broke the whole-binary
+  gate with a cascade (1,101,999 differing bytes), isolated by comm of the
+  candidate vs registry, verified per-row, dropped, re-gated green. This is
+  the merge-to-candidate safety working as designed.
+- Registry: 372 / 363. All gates green; SHA-1 stable; 229 tests.