Plan and execute statistically valid A/B and growth experiments.
npx skills add https://github.com/Victaaaw/Ledga --skill ab-test-setup-victaaaw
Please help me install this Agent Skill. Skill: ab-test-setup Source: https://github.com/Victaaaw/Ledga/tree/main/.agents/skills/ab-test-setup Command: npx skills add https://github.com/Victaaaw/Ledga --skill ab-test-setup-victaaaw
Plan and execute statistically valid A/B and growth experiments to help teams decide between alternative approaches and build a systematic experimentation program.
Define your test hypothesis, estimate the required sample size, and set up a two-variant test following the hypothesis framework.
High-intent search queries and answers about installing and using this skill.
To calculate sample size for an A/B test, you must first define a clear hypothesis, expected outcomes, and primary metrics. This approach enforces predefined sample size calculation alongside variant design to ensure statistically valid experiment results.
A/B/n tests compare multiple independent variants against a control, while multivariate tests (MVT) evaluate combinations of changes across elements. Both test types require clear hypothesis definition, traffic allocation, and guardrail metrics to ensure robust growth experimentation.
Yes, this experimentation framework supports UX and marketing split URL tests. It is applicable to product, marketing, and UX testing scenarios, allowing you to define variants, allocate traffic, and analyze results across different testing environments.
Setting up metrics for growth experiments requires defining primary metrics for main outcomes, secondary metrics for additional insights, and guardrail metrics to prevent harm. This structured metric selection ensures rigorous analysis and prevents misleading experiment results.
Stopping rules should be applied during A/B test analysis once the experiment reaches the predefined sample size and statistical significance. Enforcing these rules prevents early stopping bias and ensures rigorous, playbook-style documentation of learnings.
Veloxia-agencyDesign statistically valid A/B tests with hypotheses, metrics, and sample sizes.
OpenLaborIdentify and track A/B experiments with bootstrap confidence intervals and Mann-Whitney U tests.
delta-and-betaDesign and evaluate A/B experiments to validate campaign and feature changes.
Gull-StackDesigns, plans, and analyzes A/B tests including hypothesis, metrics, and sample size calculations.
