ML Fail-Fast Validation POC validation patterns to catch issues before committing to long-running ML experiments. Self-Evolving Skill : This skill improves through use. If instructions are wrong, parameters drifted, or a workaround was needed — fix this file immediately, don't defer. Only update for real, reproducible issues. When to Use This Skill Use this skill when: - Starting a new ML experiment that will run for hours - Validating model architecture before full training - Checking gradient flow and data pipeline integrity - Implementing POC validation checklists - Debugging prediction co…