Design and analyze A/B tests to determine treatment effects on defined metrics.
npx skills add https://github.com/arbazkhan971/godmode --skill experiment-arbazkhan971
Please help me install this Agent Skill. Skill: experiment Source: https://github.com/arbazkhan971/godmode/tree/main/skills/experiment Command: npx skills add https://github.com/arbazkhan971/godmode --skill experiment-arbazkhan971
This Skill helps teams design, run, and interpret A/B experiments with statistical rigor so they can validate product changes, avoid false positives from peeking, and protect revenue and performance guardrails.
Run an A/B test to improve signup conversion by 10% and compute required sample size, deterministic assignment, and decision rules with 80% power.
High-intent search queries and answers about installing and using this skill.
To calculate A/B test sample size and power, you define a minimum detectable effect (MDE) and apply statistical power analysis. This Skill computes the required traffic for your specified MDE and decision rules to ensure reliable experiment results.
Yes, it works with Statsig, Optimizely, and GrowthBook. The Skill detects existing experimentation SDKs and configurations, inspecting assignment logic and exposure logging to validate your web and mobile platform rollouts.
To avoid false positives from peeking during A/B testing, apply frequentist and Bayesian decision rules with strict guardrail monitoring. This Skill enforces statistical rigor and performs SRM checks to protect your revenue and performance metrics.
SRM checking, or Sample Ratio Mismatch checking, validates that traffic assignment matches expected ratios across experiment variants. It acts as a guardrail to detect logging errors or assignment bugs that could invalidate your A/B test treatment effects.
Yes, it supports feature flag ramp schedules and cleanup workflows. The Skill manages the entire experimental lifecycle, from deterministic assignment and guardrail monitoring to producing clear ship or kill verdicts for your product rollouts.
Use frequentist or Bayesian decision rules based on your experimental needs for determining treatment effects. This Skill applies both methodologies to analyze your defined metrics, enforcing latency and error-rate guardrails to output clear ship/kill decisions.
hopeovertureManage feature flags for progressive rollouts and A/B testing in web applications.
SermilionAutomate feature flag removal and legacy code cleanup after full rollout.
LauraFlorentinMonitors agent performance and generates metrics for CI/CD pipelines and audits.
ejolietGuide statistical analysis with test selection, diagnostics, and APA-style reporting.
