Data Result Validation

An honest read on whether a change, test or campaign actually worked, covering effect size, sample adequacy and the traps. Use for A/B tests, before and after comparisons or pilots.

5 stars 1 forksTesting & QA

Install

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add eshmith7/claude-skills-library --skill data-result-validation
Manual install
git clone --depth 1 https://github.com/eshmith7/claude-skills-library.git /tmp/claude-skills-library
mkdir -p ~/.claude/skills
cp -r /tmp/claude-skills-library/data-skills/data-result-validation ~/.claude/skills/data-result-validation

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Did This Actually Work

Somebody is about to tell the leadership team that a change worked. This skill is the last check before that happens. Its job is not to confirm the result. Its job is to find every reason the result might be wrong, and then say plainly what can and cannot be concluded.

## When to use this

Trigger on: A/B test results, did the campaign work, is this statistically significant, before and after comparison, pilot results, "we changed X and Y went up", "can I present this".

## Before you start

Required: the numbers, and what you did.

Ask, briefly:
1. **Was there a control group, or is this a before and after?** This decides everything below, and it is the first section of this skill for that reason.
2. Group sizes, the metric for each group, and how long it ran. Without these nothing can be assessed.
3. **How many metrics were checked.** Testing twenty almost guarantees one looks significant by chance.
4. When and why the test was stopped, and against what rule.
5. Anything else that changed in the same window, and why the change was made in the first place.

Compute in code, never in prose. Every effect size, ratio, range and sample calculation runs as Python. A significance claim worked out inside a sentence is exactly the kind of confident wrong number this skill exists to catch.

## First question, and it decides everything

**Was there a control group, or is this a before and after?**
Read the full skill on GitHub

Details

Skill name
data-result-validation
Path
data-skills/data-result-validation/SKILL.md
Repo updated
9 September 2026

FAQ

What does the Data Result Validation skill do?

An honest read on whether a change, test or campaign actually worked, covering effect size, sample adequacy and the traps. Use for A/B tests, before and after comparisons or pilots.

How do I install the Data Result Validation skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add eshmith7/claude-skills-library --skill data-result-validation. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Data Result Validation popular?

It lives in eshmith7/claude-skills-library, which has 5 GitHub stars and 1 forks.

Related skills

Content Brief

5

eshmith7/claude-skills-library

Turn a vague content request into a brief a writer can execute without a second conversation: one question, reader, intent, angle, structure, evidence required. Use for content briefs.

Writing & Content

Content Performance Audit

5

eshmith7/claude-skills-library

Turn a performance export into one decision per asset: keep, refresh, rewrite, consolidate or delete, with cannibalisation groups found first. Use for content audits or pruning.

Writing & Content

Content Prepublication Check

5

eshmith7/claude-skills-library

Check a draft before publishing: every claim registered with its source, numbers verified, headline against body, originality and compliance flags. Use before anything goes out.

Writing & Content

Edu Assessment Analysis

5

eshmith7/claude-skills-library

Turn a question-wise mark sheet into item statistics, topic and cognitive-level gaps, groups by what students missed, and a triaged reteach plan. Use for test analysis.

Testing & QA

Xunit Test Patterns

5

vnovakovits/claude-skills

Apply Gerard Meszaros's "xUnit Test Patterns — Refactoring Test Code" (2007) — the canonical catalogue of test-automation patterns, the test-double taxonomy (Dummy, Stub, Spy, Mock, Fake), test smells, fixture strategies, and result-verification patterns. Use when writing or refactoring test code; diagnosing brittle, slow, erratic, or obscure tests; choosing and naming a test double; designing fixtures (fresh vs shared, Object Mother, Test Data Builder, setup styles); structuring the four-phase test; deciding where to draw the test boundary (subcutaneous / layer / round-trip); faking out-of-pr

Testing & QA

Never Again

1

malaysherasia-ai/claude-never-again

Capture a bug or mistake as a permanent guardrail so it cannot happen twice. Use this skill whenever a bug is fixed, a test fails and is repaired, the user says "you did that again", "don't do that", "remember this", "never do X", or any correction is made that a future session would otherwise forget. Also use when the user asks to review, promote, or clean up existing lessons and hooks.

Testing & QAIn 1 list

Browse the most popular Claude skills, all testing & qa skills, or write your own with the skills generator.