Plan A/B tests with hypothesis, metrics, sample size, and duration.
npx skills add https://github.com/konglong87/superPM --skill pm-abtest
Please help me install this Agent Skill. Skill: pm-abtest Source: https://github.com/konglong87/superPM/tree/main/skills/03-growth-iteration/pm-abtest Command: npx skills add https://github.com/konglong87/superPM --skill pm-abtest
This Skill helps you validate product or feature changes by turning vague optimization ideas into a statistically grounded A/B test plan, including hypothesis, metrics, sample size, runtime, and decision rules.
docs/03-增长迭代/A-B测试/ for execution and later analysis.Ask the AI to run an A/B test planning session for your optimization idea, and it will guide you to define the hypothesis, choose metrics, estimate sample size and duration, and generate the test plan document.
High-intent search queries and answers about installing and using this skill.
To design an A/B test, you must define a clear hypothesis, establish primary and guardrail metrics, estimate the required sample size, and set test duration rules to validate the product change. This ensures causal measurement without harming user experience.
A/B test experiment design requires primary metrics to measure the target improvement, secondary metrics for broader context, and guardrail metrics to ensure the product change does not negatively impact existing user experience.
Sample size and test duration are estimated using baseline parameters and traffic assumptions. Calculating these ensures your A/B test collects sufficient data to achieve statistical significance for reliable product optimization decisions.
Yes, a structured A/B test plan document can be generated automatically. It writes the complete experimental design, including hypotheses, metrics, and runtime rules, into a designated directory for execution and later data analysis.
The A/B testing analysis framework specifies event collection guidance and a workflow covering data cleaning, descriptive statistics, hypothesis testing, effect size calculation, and segmentation to interpret experiment results accurately.
Guardrail metrics should be used in every product optimization experiment to monitor predefined thresholds. They prevent negative impacts on user experience or core business metrics when testing new feature changes.
sicktasticPlan statistically sound A/B tests with sample size and duration calculations.
Lord1EgyptPerform statistical hypothesis testing with assumption checks and APA-style reporting.
GrupoUSWrite and execute Python scripts for data analysis, modeling, and visualization.
amanhsnDesign experiments, build predictive models, and perform causal analysis with Python, R, and SQL.