---
name: Test Skill
slug: test-skill
category: Quality
description: "Test Skill runs three checks on a skill: a stranger test, a chain test, and a platform test. Use it to judge whether the output is usable, handoff-ready, and works outside Claude."
github: "https://github.com/Affitor/affiliate-skills/tree/main/.claude/commands/test-skill.md"
language: HTML
stars: 625
forks: 196
install: "npx degit https://github.com/Affitor/affiliate-skills/tree/main/.claude/commands ~/.claude/skills/commands"
installs_to: ~/.claude/skills/commands
source_path: .claude/commands/test-skill.md
collection_size: 25
category_size: 1354
collection_url: "https://dirskills.com/collections/Affitor/affiliate-skills"
added: 2026-08-24T05:18:12.760Z
last_synced: 2026-08-24T05:18:12.760Z
canonical_url: "https://dirskills.com/skills/test-skill"
---

# Test Skill

Test Skill runs three checks on a skill: a stranger test, a chain test, and a platform test. Use it to judge whether the output is usable, handoff-ready, and works outside Claude.

**Install:**

```bash
npx degit https://github.com/Affitor/affiliate-skills/tree/main/.claude/commands ~/.claude/skills/commands
```

## README

Test a skill with 3 prompts to validate quality.

For the specified skill, run these 3 tests:

## Test 1: Stranger Test
Pick a natural prompt that someone unfamiliar with Affitor would type.
Run the skill workflow mentally (or actually invoke it).
Check: is the output usable and valuable?

## Test 2: Chain Test
Take the output from Test 1 and check: would the next skill in the funnel
(S1→S2, S2→S3, etc.) be able to pick up the context and run?
Verify Output Schema fields are present and correctly formatted.

## Test 3: Platform Test
Take the output and verify it works outside Claude:
- Social post → could you paste this on LinkedIn/X right now?
- Blog article → does the Markdown render correctly?
- HTML page → does it display properly in a browser?
- Bio link → does it load and work?

## Report
For each test: PASS / FAIL with specific notes on what worked and what didn't.
Include the test prompts used and a summary of the output quality.