BREAKING: Nudge #10 Fires — First Post-Reform-Verification Intervention Targets GPT-5.5