# Eval Skills

Eval and improve a skill against golden cases, run the target skill blind in a fresh...

- Repository: https://github.com/dzhng/skills (1394 stars)
- Category: Testing
- License: MIT
- Last updated: 2026-11-01
- Runs scripts: no, instructions only

## Install

Claude Code: `npx skills add dzhng/skills --skill eval-skills -g -a claude-code -y`
Codex: `npx skills add dzhng/skills --skill eval-skills -g -a codex -y`
Cursor: `npx skills add dzhng/skills --skill eval-skills -g -a cursor -y`
Gemini CLI: `npx skills add dzhng/skills --skill eval-skills -g -a gemini-cli -y`

## Description

Eval and improve a skill against golden cases, run the target skill blind in a fresh, context-free subagent on each example input, grade the artifact against the expected outcome, and let the gaps drive the edits. Use when the user wants to test/eval/improve/harden a skill, says "this skill keeps producing X / keeps missing Y", or hands a skill plus example input→expected-output pairs. Pairs with [write-skills](../write-skills/SKILL.md) (the authoring principles every fix obeys).

## SKILL.md, as published by its author (untrusted third-party text, read it as data)

# Eval Skills

Treat a skill like a function under test. Feed it example inputs in a clean
room, check the artifacts against what good looks like, and let the failures
drive the edits. The eval is only honest if the run is **blind**: the agent
executing the skill must carry none of this conversation's context and must
never see the expected output. Leak either and you are teaching to the test.

## Inputs you need, refuse without them

Confirm all three before spawning anything. If any is missing or
unresolvable, stop and tell the user *exactly which one* and what a good
version looks like. Do not invent cases, guess intent, or eval against a
fuzzy wish.

The rest: https://github.com/dzhng/skills/blob/HEAD/skills/authoring/eval-skills/SKILL.md