← All skills

eval

100 / 100 · Low risk

Fine to try.

By drsh4dow/pi-setup (GitHub) · other · 10 stars · scanned at commit f793659

Measure how a skill, prompt, system-prompt, or agent-file change affects agent behavior before promoting it: blinded candidates on an organic task, a hidden rubric, one blinded judge, and a check of each candidate's transcript. Use for 'eval this change', or before merging a change whose only test is how agents behave.

Works with any agent that supports SKILL.md skills.

What this skill can do

We found nothing in the files we read that matches our rules. That does not prove the skill is safe: see the limits of static scanning.

Install

Paste this into your terminal. It downloads exactly the version we scanned, not whatever is newest.

if [ -e ~/.agents/skills/'eval' ] || [ -L ~/.agents/skills/'eval' ]; then echo 'A skill folder named eval already exists — not replacing it' >&2; false; else d=$(mktemp -d) && git clone --quiet --filter=blob:none --no-checkout 'https://github.com/drsh4dow/pi-setup' "$d" && git -C "$d" checkout --quiet 'f7936596c1ba2e1472dadccc3f111b9d517fa96c' -- 'agent/skills/eval' && mkdir -p ~/.agents/skills && cp -R "$d"/'agent/skills/eval' ~/.agents/skills/'eval' && rm -rf "$d"; fi

Codex reads ~/.agents/skills directly. For Claude Code, Cursor, Copilot, Gemini CLI, Antigravity or OpenCode, also link it. For Claude Code: ln -s ~/.agents/skills/'eval' ~/.claude/skills/'eval' (other agents: ~/.cursor/skills, ~/.copilot/skills, ~/.gemini/skills, ~/.gemini/config/skills, ~/.config/opencode/skills). Or install through the finder skill, which does both.

Badge for your README

The badge for drsh4dow/pi-setup shows the score of its lowest-scoring skill.

[![SkillTrust](https://zalavijaysinh.github.io/skilltrust/badges/drsh4dow/pi-setup.svg)](https://zalavijaysinh.github.io/skilltrust/repos/drsh4dow/pi-setup/)

Score history