| name | cumcm-data-quality-audit |
| description | Audit CUMCM datasets for missing values, units, duplicates, outliers, inconsistent identifiers, impossible ranges, time/order problems, leakage, and preprocessing decisions. Use when reviewing raw or processed data before modeling, writing data cleaning notes, or deciding whether results are trustworthy. |
CUMCM Data Quality Audit
Use this skill before fitting models or when a result looks suspicious.
Workflow
- Inventory files, fields, row counts, units, time span, and identifiers.
- Check missingness, duplicates, impossible values, outliers, and inconsistent encodings.
- Separate true data problems from domain-valid rare cases.
- Record every cleaning decision and whether it affects reproducibility.
- Recommend plots/tables needed to justify preprocessing in the paper.
- Read
references/checklist.md for the audit list.
- Read
references/issue-taxonomy.md when classifying data findings.
- Read
references/examples.md when writing the data-quality section.
Guardrails
- Do not silently drop rows or columns.
- Do not replace missing data without recording the rule.
- Keep raw data immutable; write cleaned data to a separate processed path.