SkillEvaluator
Multi-tier framework for evaluating AI agent skills and plugins with quality gates, semantic overlap detection, synthetic evaluation dataset generation, and live agent evaluation that measures how...
Install
npx skills add NVIDIA/SkillEvaluator -a claude-code
npx skills add NVIDIA/SkillEvaluator -a claude-code -g
Or copy the skill folder into .claude/skills/ (project) or ~/.claude/skills/ (all projects).
npx skills add NVIDIA/SkillEvaluator -a codex
npx skills add NVIDIA/SkillEvaluator -a codex -g
Or copy the skill folder into .agents/skills/ (project) or ~/.agents/skills/ (all projects).
See the repository README for installation instructions.
When to use
Use it before you publish or adopt a skill, to find out whether it actually improves results and doesn't overlap with skills you already have.
SkillEvaluator is NVIDIA's framework for evaluating agent skills and plugins. It checks quality, security, overlap and compatibility, and measures a skill's effect on task results. The process starts with quality checks, then compares runs with and without the skill. Features include quality gates, semantic overlap detection, synthetic evaluation dataset generation and live agent evaluation.
It's a Python CLI on PyPI for Python 3.12 to 3.14, part of NVIDIA's Verified Skills pipeline, and has a paper, docs and a blog post. NVIDIA's research it draws on found that how a skill reads and how it performs can be unrelated.
Related skills
-
rea
Lets your agent reverse engineer an app from the outside, down to native binaries, and then build its own version.
-
superpowers
A set of development habits for your agent: test first, debug systematically, and plan before you code.
-
skills
The skills Matt Pocock uses in his own day-to-day engineering work.
-
ECC
A ready-made setup for coding agents that adds skills, persistent memory and security checks.