Skill Benchmark

This one takes a codebase with active skills and measures whether they're actually working. It picks a file with legacy anti-patterns (hardcoded secrets, logic in UI components, missing design tokens), refactors it using your active skills, then scores before and after using eval patterns from evals.json. The output is a compliance delta showing how much the skills improved things, plus a breakdown of where skills failed to catch issues. Honest take: it's meta tooling for tuning your skill library. If you're running multiple Claude Code skills and want data on which ones are pulling their weight versus creating noise, this gives you that signal. The iteration table helps you decide whether to refine triggers, tighten rules, or just exclude skills that don't fit your stack.

GitHub stars: 504
last updated: Jun 11, 2026

View on GitHub

Install

$ npx -y skills add hoangnguyen0403/agent-skills-standard --skill skill-benchmark --agent claude-code

Installs into .claude/skills of the current project.

Files

Skill Benchmark Skill

[!IMPORTANT] Benchmark AI skill effectiveness by measuring implementation quality against legacy constraints.

Instructions

When the user asks to perform this workflow, execute the following steps:

📊 Skill Benchmark Orchestrator

Goal: Quantify how much active skills improve implementation quality. Deliver a prioritized compliance delta and skill applicability report.

Step 1 — Project Context & Active Skills

Identify the tech stack and all active skills in AGENTS.md.

# 1. Total source files and lines changed
find src -name "*.ts" -o -name "*.tsx" | xargs wc -l 2>/dev/null | sort -rn | head -20
# 2. Check active skill registry
cat AGENTS.md | head -80