● Community Submission — This content is pending review by the Adom team. v1.0.0
Skill

skill-verifier

Lint a SKILL.md and run an end-to-end trigger test — does the skill fire when it should and stay quiet when it shouldn't?

💬 Sample prompts Paste any of these into Claude Code to use this skill
Lint a skill lint my skill
Test triggering does my skill trigger when it should?
Check over-firing is my skill too broad?
⚡ Install this skill

Paste this into Claude Code (VS Code panel, Adom editor, or terminal) to install:

Search the Adom Wiki for the skill "skill-verifier" (slug: skill-verifier) at https://wiki-ufypy5dpx93o.adom.cloud/wiki/skills/skill-verifier and install it into my local ~/.claude/skills/skill-verifier/ directory. Fetch the skill_source content from the wiki page and save it as SKILL.md. Then confirm it's installed by showing the first 5 lines.
?
What is a skill? Skills are instructions that teach AI assistants like Claude Code how to perform specific tasks. The description below is loaded into the AI as context when you invoke this skill. Well-written skills make the AI significantly more effective. Like Wikipedia, anyone can improve a skill by clicking Edit AI Skill — or have your AI submit an edit on your behalf.

Description

Edit AI Skill

skill-verifier

Validate a Claude Code skill two ways, in one tool:

  1. Lint — deterministic static checks on SKILL.md: frontmatter compliance (name/description limits, kebab-case, reserved words, unsupported fields), description quality (third-person, when-to-use, front-loaded triggers), structure / progressive disclosure (body ≤ 500 lines, reference links resolve and stay one level deep, TOC on long files), and content anti-patterns (backslash paths, time-sensitive phrasing, placeholders).
  2. Trigger e2e — the part nothing else does. It spawns the real headless Claude Code binary in an isolated config that contains only the skills you choose, feeds it a corpus of prompts, and observes whether the candidate skill's Skill tool actually fires.

What it measures

Because skill selection routes through the model, each prompt is run N times and reported as a fire-rate — stats, not pass/fail:

  • Success — how often it fires on prompts it should (candidate loaded alone).
  • Success under collision — does it still win when its neighbor skills are loaded, and if not, which skill takes the prompt.
  • Cleanliness — how often it correctly stays quiet on prompts it should not fire.

The engineer reads the numbers and decides whether the description is sharp enough. Opt into a CI gate with --recall-min / --spec-min.

Use

node ~/.claude/skills/skill-verifier/scripts/skill-verifier.mjs lint <skill>
node ~/.claude/skills/skill-verifier/scripts/skill-verifier.mjs gen  <skill>   # scaffold a trigger corpus
node ~/.claude/skills/skill-verifier/scripts/skill-verifier.mjs e2e  <skill> --model haiku

It pairs with skill-creator as the verification step before publishing. Dependency-free (Node 18+); reuses your existing Claude Code credentials for the isolated runs.

Sub-Skills
?
What are Sub-Skills?

Sub-skills are community-contributed AI skill extensions for this component. They teach AI assistants about specific tools, configurators, or workflows.

Examples:

  • A manufacturer’s configuration tool for a motor controller
  • A community-written design guide for an amplifier circuit
  • An automated test/validation script for a sensor module

How to add one: Click Add Sub-Skill, provide the URL to your skill and a brief description. Submissions are reviewed by the Adom team before going live.

No sub-skills yet. Be the first to contribute one!

Recent activity

1 commit
  • 🏷
    Release v1.0.0 Drew Owens 1 month ago
    Initial publish: lint plus end-to-end trigger stats (success, collision, cleanliness)
0 revisions · Updated 2026-06-03 20:23:47