data-validation

Pre-delivery QA checklist, common data analysis pitfalls, result sanity checking, and documentation standards for reproducibility.

Run through this checklist before sharing any analysis with stakeholders.

The problem: A many-to-many join silently multiplies rows, inflating counts and sums.

在與利害關係人分享之前對分析進行品質檢查——方法檢查、準確性驗證和偏差檢測。在檢查錯誤分析、檢查生存偏差、驗證聚合邏輯或準備可重複性文件時使用。 來源：anthropics/knowledge-work-plugins。