Automates the setup of A/B tests and experimental frameworks when you need to validate product decisions with real data. It generates experiment configs, helps you define success metrics, and sets up the statistical scaffolding so you're not manually calculating sample sizes or significance thresholds. Useful when you're moving beyond gut decisions but don't want to context switch into statistics mode. The structured approach keeps you honest about what you're actually testing and when to call results, which is half the battle with experiments that drag on without clear stopping points.
npx skills add https://github.com/alirezarezvani/claude-skills --skill experiment-designer