This is a structured pre-flight checklist for data analysis, designed to catch the embarrassing errors before they reach stakeholders. It walks through methodology review, calculation spot-checks, and a catalog of common pitfalls like join explosions and survivorship bias. The output is a three-tier confidence rating (ready to share, share with caveats, needs revision) plus specific findings. Use it when you've finished an analysis and want a systematic second look before the presentation. The checklist is thorough enough to be genuinely useful but still requires you to actually verify the numbers, it's not magic validation.
npx -y skills add anthropics/knowledge-work-plugins --skill validate-data --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
juliusbrussee/caveman
mattpocock/skills
shadcn/improve
obra/superpowers
forrestchang/andrej-karpathy-skills
vercel-labs/skills