# Ab Test Designer

> |

- **Type:** Skill
- **Install:** `agentstack add skill-lionelsimai-claude-skills-collection-ab-test-designer`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [lionelsimai](https://agentstack.voostack.com/s/lionelsimai)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [lionelsimai](https://github.com/lionelsimai)
- **Source:** https://github.com/lionelsimai/claude-skills-collection/tree/main/skills/ab-test-designer

## Install

```sh
agentstack add skill-lionelsimai-claude-skills-collection-ab-test-designer
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# A/B Test Designer

## Overview
Designs rigorous A/B tests with clear hypotheses, variant specifications, sample size calculations, and analysis plans.

## Workflow

### Step 1: Define the Test
1. **What are you testing?** (page, email, ad, feature)
2. **Current metric**: What's the baseline performance?
3. **Goal**: What improvement are you hoping for?
4. **Traffic/volume**: How many users/emails/impressions per day?

### Step 2: Structure the Test

## Output Format

```markdown
# A/B Test: [Test Name]

## Hypothesis
If we [change X], then [metric Y] will [increase/decrease] by [Z%] because [reasoning].

## Test Details
- **Type**: A/B / A/B/C / Multivariate
- **Primary metric**: [what you're measuring]
- **Secondary metrics**: [supporting metrics]
- **Guardrail metrics**: [what shouldn't get worse]

## Variants

### Control (A)
[Description of current experience]

### Variant (B)
[Description of the change]
[Mockup/wireframe description if applicable]

## Sample Size & Duration
- **Baseline conversion**: [X%]
- **Minimum detectable effect**: [X%]
- **Statistical significance**: 95%
- **Required sample size**: [N per variant]
- **Estimated duration**: [X days]

## Analysis Plan
1. Wait for minimum sample size before checking
2. Check primary metric first
3. Segment analysis: [segments to check]
4. If significant: [implementation plan]
5. If not significant: [next steps]

## Risks & Considerations
- [Risk 1]: [mitigation]
- [Risk 2]: [mitigation]
```

## Quality Checklist
- [ ] Hypothesis is specific and falsifiable
- [ ] One primary metric defined
- [ ] Sample size calculated
- [ ] Duration estimated
- [ ] Analysis plan prevents peeking bias
- [ ] Guardrail metrics identified

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [lionelsimai](https://github.com/lionelsimai)
- **Source:** [lionelsimai/claude-skills-collection](https://github.com/lionelsimai/claude-skills-collection)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-lionelsimai-claude-skills-collection-ab-test-designer
- Seller: https://agentstack.voostack.com/s/lionelsimai
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.