What problem does it solve? Messy tabular data from CSVs, spreadsheet pastes, and system exports often contains inconsistent formats, duplicates, and ambiguous values that get silently mangled by ad-hoc cleanup. This Skill enforces a disciplined workflow that preserves the original data, surfaces ambiguity instead of guessing, and accounts for every row. ## Core Features & Use Cases - Source Preservation and Profiling: Records row counts, column types, missing values, and anomalies before any edits, never overwriting the original input. - Confirmed Transformation Rules: Proposes target formats, duplicate keys, and missing-value policies, requiring confirmation before any meaning-changing merge, deletion, or imputation. - Auditable Output: Delivers a cleaned dataset plus a transformation log with quarantined rows, dropped-row reasons, and an explicit row equation (input = output + dropped + quarantined). - Use Case: You receive a 5,000-row customer export with mixed date formats like 02/03/24, inconsistent casing, and possible duplicates. The Skill profiles the file, proposes normalization rules, quarantines ambiguous rows for your review, and produces a cleaned file with a full audit trail. ## Quick Start Clean and standardize the attached messy CSV export, deduplicate it, and show me a transformation log accounting for every row.
