Data Analysis Notebook Skill
Reproducible, honest analysis: TL;DR-first notebooks, costed filters, join guards, and uncertainty by default.
by DataForge Labs5.0(1)428 installs
About this skill
What's inside
SKILL.mdwith the 5-cell notebook structure (TL;DR first, validated load, one question per cell, limitations last)- Data hygiene rules: costed filters, fan-out-guarded joins, explicit missing-data policy
- Statistical honesty rules and takeaway-titled chart standards
references/pandas-patterns.md— vetted snippets: validation loads, bootstrap CIs, timezone-safe resampling
Use case
Data-analysis skills are a staple of official example collections. This one encodes the discipline that separates "ran some pandas" from analysis a decision-maker can trust.
Versions
v1.2.0Add bootstrap CI pattern; strengthen join validation guidance7/20/2026Reviews
5.0(1)validate='many_to_one' everywhere nowToolsmith Collective · 7/20/2026
The join fan-out guard found a silent row-duplication bug in a dashboard we'd trusted for months.