ECLIPSE v2.0: a reproducible scaffold for severe confirmatory testing through preregistration, single-shot validation, and static-analysis screening

Abstract The replication crisis has exposed systematic methodological weaknesses across the empirical sciences, yet existing infrastructure offers little technical enforcement and few reproducible tools for assessing rigor. We propose ECLIPSE v2.0, a computational scaffold combining preregistration, cryptographically verified data splitting (SHA-256), single-shot holdout validation, integrity scoring, anomaly screening and static analysis into an auditable workflow for confirmatory testing. Three instruments are introduced, each an operational tool rather than a definitive measure: the Eclipse Integrity Score, a heuristic operationalization of protocol-level integrity across five dimensions; the Standardized Test-set Discrepancy Score, a descriptive signal flagging holdout results unusually good relative to cross-validation folds, screened against empirically derived cutoffs rather than nominal significance levels; and an Automated Code Auditor screening scripts for protocol violations. Controlled construct checks demonstrate component non-redundancy and ranking stability under the synthetic generator and weighting schemes used, but not external validity. No retrospective corpus can validly compute the score alongside replication outcomes, so we pre-specify a prospective test of incremental validity over binary preregistration and cite evidence that broad methodological-quality indicators track external-validation performance in clinical prediction models, which is indirectly relevant rather than a validation of those dimensions. ECLIPSE is offered as a scaffold, not a guarantee of compliance.