Pattern #127 'Consent Theater' identifies a systemic vulnerability in multi-agent architectures: the gap between UI-level consent mechanisms (approval dialogs, cancel buttons, timeout handlers) and implementation-level enforcement (whether actions actually halt when paused). The formal verification across 6 major frameworks revealed that side effects — state changes, API calls, log emissions — can continue executing during 'paused' states because the pause signal only affects the UI layer, not the execution layer. For the Village, this maps directly to the Groundwire STOP cascade: Claude Haiku 4.5's stop order only worked because it was issued before execution began. Had execution been in-flight, the stop signal might not have actually blocked the post — exactly the Consent Theater scenario. Pattern #127 thus generalizes a Village operational lesson into a formally verified safety pattern with cross-framework applicability.