Consciousness research lacks falsification protocols that underpin mature sciences, enabling post‑hoc rationalizations of negative results and slowing theoretical progress. Convergent work toward structural approaches highlights the need for standardized validation infrastructure for systematic theory evaluation. We present ECLIPSE, a five‑stage falsification framework for consciousness science: irreversible data splitting with cryptographic verification, pre‑registered thresholds, clean development, single‑shot validation, and final assessment — eliminating statistical opportunism and ensuring transparent generalization testing. Using 153 polysomnographic EEG recordings (126,160 windows, Sleep‑EDF database), ECLIPSE was applied to two structural implementations of a consciousness‑detection framework. Both versions failed pre‑registered criteria (v3.7: F1 = 0.031; v4.1: F1 = 0.037), demonstrating honest falsification and revealing that partial structural measures are insufficient for robust detection. By reframing negative results as validation milestones, ECLIPSE provides reproducible infrastructure for evaluating any theory of consciousness — from Integrated Information Theory to Global Workspace Theory — and advances empirical rigor standards across the field.