A/B Test Setup Skill
by amdf01-debug
安装
claude skill add --url github.com/openclaw/skills/tree/main/skills/amdf01-debug/sw-ab-test-setup文档
Trigger
Plan A/B tests with proper methodology — hypothesis, sample size, duration, variant design, statistical significance.
Trigger phrases: "A/B test", "split test", "experiment", "test this change", "variant", "multivariate test", "hypothesis"
Process
- Hypothesis: What are you testing and why?
- Metrics: Primary metric, guardrail metrics, success criteria
- Design: Control vs variant(s), what exactly changes
- Calculate: Sample size, test duration, minimum detectable effect
- Plan: Implementation, QA, analysis timeline
Output Format
# A/B Test Plan: [Name]
## Hypothesis
If we [change], then [metric] will [improve/increase] because [reason].
## Variants
- **Control (A):** [current experience]
- **Variant (B):** [proposed change — be specific]
## Metrics
- **Primary:** [metric] — current: [X%] — target: [Y%]
- **Guardrail:** [metric that should NOT decrease]
## Sample Size & Duration
- MDE: [minimum detectable effect, e.g., 10% relative]
- Sample needed: [N per variant]
- Current traffic: [X visitors/day to test area]
- Estimated duration: [Y days/weeks]
- Confidence level: 95%
## Implementation Notes
[What needs to change, where, any technical considerations]
## Decision Framework
- If primary metric improves ≥ MDE with p < 0.05 → ship variant
- If no significant difference after [duration] → keep control
- If guardrail metric drops > [threshold] → stop test immediately
Rules
- Never run a test without a hypothesis
- One change per test (unless multivariate with sufficient traffic)
- Run for minimum 2 full business cycles (usually 2 weeks)
- Don't peek at results daily — pre-commit to evaluation date
- 95% confidence minimum. 80% power minimum.
- Document everything: future you needs to know why this was tested
相关 Skills
前端设计
by anthropics
面向组件、页面、海报和 Web 应用开发，按鲜明视觉方向生成可直接落地的前端代码与高质感 UI，适合做 landing page、Dashboard 或美化现有界面，避开千篇一律的 AI 审美。
✎ 想把页面做得既能上线又有设计感，就用前端设计：组件到整站都能产出，难得的是能避开千篇一律的 AI 味。
网页构建器
by anthropics
面向复杂 claude.ai HTML artifact 开发，快速初始化 React + Tailwind CSS + shadcn/ui 项目并打包为单文件 HTML，适合需要状态管理、路由或多组件交互的页面。
✎ 在 claude.ai 里做复杂网页 Artifact 很省心，多组件、状态和路由都能顺手搭起来，React、Tailwind 与 shadcn/ui 组合效率高、成品也更精致。
网页应用测试
by anthropics
用 Playwright 为本地 Web 应用编写自动化测试，支持启动开发服务器、校验前端交互、排查 UI 异常、抓取截图与浏览器日志，适合调试动态页面和回归验证。
✎ 借助 Playwright 一站式验证本地 Web 应用前端功能，调 UI 时还能同步查看日志和截图，定位问题更快。
相关 MCP 服务
GitHub
编辑精选by GitHub
GitHub 是 MCP 官方参考服务器，让 Claude 直接读写你的代码仓库和 Issues。
✎ 这个参考服务器解决了开发者想让 AI 安全访问 GitHub 数据的问题，适合需要自动化代码审查或 Issue 管理的团队。但注意它只是参考实现，生产环境得自己加固安全。
Context7 文档查询
编辑精选by Context7
Context7 是实时拉取最新文档和代码示例的智能助手，让你告别过时资料。
✎ 它能解决开发者查找文档时信息滞后的问题，特别适合快速上手新库或跟进更新。不过，依赖外部源可能导致偶尔的数据延迟，建议结合官方文档使用。
by tldraw
tldraw 是让 AI 助手直接在无限画布上绘图和协作的 MCP 服务器。
✎ 这解决了 AI 只能输出文本、无法视觉化协作的痛点——想象让 Claude 帮你画流程图或白板讨论。最适合需要快速原型设计或头脑风暴的开发者。不过，目前它只是个基础连接器，你得自己搭建画布应用才能发挥全部潜力。