plan: add gpt55 cfold review phase

cc-ci-plan/JOURNAL.md

@@ -578,3 +578,16 @@ session cc-ci-orchestrator-stale can be killed; recipe-mirrors org still private
 - Verified live tmux mapping: `cc-ci-builder`, `cc-ci-adv`, and `cc-ci-orchestrator` are all `opencode`;
   `cc-ci-watchdog` is running. The watchdog will hourly-wake `cc-ci-orchestrator` via the existing
   `ORCH_WAKE_INTERVAL=3600` path and will apply the existing limit-window nudge/restart handling.
+
+## 2026-06-12 ~16:05 — queued GPT-5.5 post-cfold no-loss review
+- Operator requested one additional pass at the end of `cfold`: GPT-5.5 reviews the cfold
+  implementation and confirms no custom test/coverage/assertion/fixture behavior was lost.
+- Added `plan-phase-cf55-gpt55-cfold-review.md` and inserted `cf55` immediately after `cfold`, before
+  `pvfix`. Persisted queue is now
+  `rcust;shot;lvl5;bsky;dstamp;mailu;kuma;mailu;drone;cfold;cf55;pvfix;pvcheck;ghost`.
+- Set per-phase model overrides in `.cc-ci-logs`: `.loop-model-cf55=openai/gpt-5.5` and
+  `.loop-model-adv-cf55=openai/gpt-5.5`. Current `cfold` loops stay on OpenCode GPT-5.4; when cfold
+  writes real `## DONE`, watchdog should auto-transition to `cf55` and start both loops on GPT-5.5.
+- Bounced only `cc-ci-watchdog` so it loaded the 14-phase queue without interrupting the active builder
+  and adversary sessions. Verified `launch.sh status`: `cfold [10/14]`, builder/adv/watchdog RUNNING;
+  watched sessions are still `opencode` for builder/adv/orchestrator.