compare-models-blindly

Creates seeded blinded model-comparison packets and evaluates candidate outputs twice with reversed order to control identity and presentation bias. Use when two models or prompts produce different valid answers that require qualitative judging.