技能 webask-ab-significance

📦

webask-ab-significance

Name: webask-ab-significance
Author: webaskio

內容修訂 r1 安全 📁 檔案系統存取⚙️ 外部命令

Compare WebAsk A/B Survey Results

Small survey samples can make random differences look like meaningful improvements. This skill guides comparisons using outcome counts, percentage-point differences, and checks for comparable groups.

支援: Claude Codex Code(CC)

🥉 77 銅級

Maintain this Skill? Claim it 🥉 77 銅級

使用我的智慧代理安裝

將此請求複製給你的智慧代理，其中包含權威 Skill 頁面和清單。

智能體請求

Review the Skillstore skill "webask-ab-significance" from https://skillstore.io/skills/webaskio-webask-ab-significance.md and its manifest at https://skillstore.io/api/skills/webaskio-webask-ab-significance/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

你的智慧代理仍應顯示其計畫，並依安全政策請求必要確認。

Agent 可讀資源

當 AI Agent、爬蟲或腳本需要乾淨脈絡、而不是讀取完整頁面時，請使用這些連結。

Markdown 詳情 GET /skills/webaskio-webask-ab-significance.md 已簽名 manifest GET /api/skills/webaskio-webask-ab-significance/manifest 已簽名 lockfile GET /api/skills/webaskio-webask-ab-significance/lockfile

測試它

正在使用「webask-ab-significance」。 Compare contact submissions: A has 42 of 180, and B has 34 of 179. Both groups use the same period and source.

預期結果:

Variant A: 42 of 180 submitted contact details, or 23.3%.
Variant B: 34 of 179 submitted contact details, or 19.0%.
Observed difference: 4.3 percentage points in favor of A.
These counts alone do not establish a reliable winner. A formal test and sample-size plan are needed before claiming significance.

正在使用「webask-ab-significance」。 A completed 60 of 100 surveys last month from email traffic. B completed 75 of 100 this month from paid traffic.

預期結果:

Variant A: 60% completion. Variant B: 75% completion.
Observed difference: 15 percentage points in favor of B.
The periods and traffic sources differ, so the comparison cannot isolate the variant effect.
Compare groups from the same period and traffic source before choosing a winner.

安全審計

安全

v1 • 2026/9/29 開啟版本化報告

Both static findings are false positives: README links provide repository navigation, and SKILL.md backticks format MCP tool names. No evidence found of prompt injection, unauthorized execution, or data exfiltration in the reviewed files. External MCP implementation behavior is outside this review.

已掃描檔案

119

分析行數

複核項目

已忽略的誤報

風險因素

📁 檔案系統存取 (1)

README.md:1

⚙️ 外部命令 (1)

SKILL.md:15

最近一次完成的靜態與語意稽核未偵測到已確認的安全發現。這並不能證明該 Skill 沒有副作用。

審計單位: codex

分享與引用此報告

分享版本化評估報告、中性徽章、嵌入卡片與引用。Skillstore 只報告證據，不判斷該 Skill 是否安全。

開啟版本化報告

複製報告連結

https://skillstore.io/skills/webaskio-webask-ab-significance/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 徽章

[![Skillstore security assessment](https://skillstore.io/badges/skills/webaskio-webask-ab-significance/security.svg)](https://skillstore.io/skills/webaskio-webask-ab-significance?utm_source=security_passport_badge)

HTML 徽章

<a href="https://skillstore.io/skills/webaskio-webask-ab-significance?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/webaskio-webask-ab-significance/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

嵌入卡片

<iframe src="https://skillstore.io/embed/skills/webaskio-webask-ab-significance.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

學術引用 (APA · BibTeX · CFF)

APA 引用

webaskio. (2026). webask-ab-significance security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/webaskio-webask-ab-significance/audits/1

BibTeX 引用

@techreport{webaskio-webaskio-webask-ab-significance-2026,
  author = {webaskio},
  title = {webask-ab-significance security audit report (audit version 1)},
  institution = {Skillstore},
  year = {2026},
  number = {1},
  url = {https://skillstore.io/skills/webaskio-webask-ab-significance/audits/1},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "webask-ab-significance security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
  - name: "webaskio"
date-released: "2026-09-29"
url: "https://skillstore.io/skills/webaskio-webask-ab-significance/audits/1"
identifiers:
  - type: other
    value: "skillstore:webaskio-webask-ab-significance:audit:1"
    description: "Skillstore immutable audit report identifier"

Skillstore 評分

評分依據證據置信度: 中

架構

可維護性

內容

社群

規範符合度

你可以打造什麼

Compare survey completion

Compare completion counts and rates for two survey versions while checking observation periods and traffic sources.

Review contact collection

Compare contact submission rates across survey variants without selecting a winner from a small observed difference.

Assess experiment readiness

Review whether response counts and group comparability support a decision or require more data.

試試這些提示

Compare two variants

Compare variants A and B for WebAsk survey [survey]. Use survey completion as the success metric.

Report conversion differences

Compare contact submissions for variants [A] and [B] in survey [survey] during [period]. Show counts, denominators, rates, and percentage-point differences.

Check group comparability

Review variants [A] and [B] for survey [survey]. Check periods, traffic sources, and sample sizes before interpreting [metric]. Identify missing information.

Review decision uncertainty

Assess whether survey [survey] supports choosing variant [A] or [B] for [metric]. Separate observed differences from statistical evidence. Explain sample-size assumptions and missing inputs.

最佳實務

Define the success metric and denominator before reviewing variant results.
Compare groups from the same period and traffic source, and retain the underlying counts.
Use a prespecified statistical test and power calculation to validate decisions beyond the skill's sample-size heuristics.

避免

Declaring a winner from a small percentage difference without considering sample size.
Comparing variants from different periods or traffic sources as if the groups were equivalent.
Repeatedly checking results and stopping as soon as the preferred variant leads.

常見問題

What data does the skill need?

It needs variant exposure counts, completion counts, the chosen success metric, observation periods, and traffic sources.

Which tools does it use?

It names get_quiz_summary and get_quiz_report as external WebAsk MCP dependencies for retrieving survey data.

Does it calculate formal statistical significance?

The files provide comparison instructions and sample-size heuristics. They do not specify a formal test, significance threshold, or confidence-interval method.

Can it determine the exact sample size required?

It asks for a collection estimate but supplies only rough guidance. Exact planning requires assumptions about baseline rate, detectable effect, significance, and power.

Can it modify surveys or require an upgrade?

The documented tools read reports and summaries. The instructions prohibit upgrade suggestions and require stopping when service limits prevent progress.

Which assistants and languages are supported?

The listing supports Claude, Codex, and Claude Code. The skill instructs the assistant to respond in the user language.