KArtSell.Aegis/docs/templates/MODEL_EVALUATION_EVIDENCE.md

Model Evaluation Evidence

• Evaluation ID:
• Operation Code / Schedule Version:
• Scope / Cohort / Window:
• Dataset ID / Hash:
• Model / Config / Code / Contract Version:
• Metric Definition Version:
• PIT cutoff / Published-at rule:
• Sample size / Missing / Quarantine:
• Results and confidence intervals:
• Baseline / Champion / Challenger:
• Costs / Tax / FX / Slippage scenario:
• Counter-evidence:
• Operational integrity errors:
• Gate result and blocking reasons:
• Source / Assumption / Unknown / Decision Required:
• Commands / Artifact hashes:
• Independent validator:
• Maker / Checker: