diff --git a/phase-ends/CURRENT_PHASE.md b/phase-ends/CURRENT_PHASE.md
index b5de50c..e539765 100644
--- a/phase-ends/CURRENT_PHASE.md
+++ b/phase-ends/CURRENT_PHASE.md
@@ -245,3 +245,17 @@ then (46-trapping exclusion count).
 **Deliverable for P9-T8:** cookbook finding 26 updated with the gap census
 (942 vs 661), the direct probe evidence, and the 0x8001FDA0/0x8001F4F4
 cluster addresses as the class's strongest members.
+
+## Cycle 3 — worker B re-engagement and coordinator absorption
+
+- Worker B resumed at 27% context (lowest of the four sessions) after a
+  self-correction: its "budget low" flag was fatigue, not measurement. Its
+  commitment: re-flag only on measured criteria (0 matches over two
+  consecutive cycles, attempts-per-match), quoting the intercom context figure.
+- Worker B re-chartered on partition A (5 call-shape rows first, then fresh
+  tier-1 leaves), with pre-read corrections (0x80101C2C/0x80102FA4 now
+  registered by the coordinator's absorption; 0x800FCC30/0x8002622C are
+  worker-A negatives).
+- Coordinator absorption continues on partitions B and C: 6 bounded
+  negatives recorded (no-frame-call library-asm family, cross-join stores,
+  alloc-tiebreak leaves) and 4 matches closed since the last ledger entry.