eval-skills

Installation
SKILL.md

eval-skills

Behavioral evaluation for Embry OS skills. Fills the gap between sanity.sh ("does it run?") and skills-ci ("does it follow rules?") by answering "does it produce correct output?".

Usage

# Evaluate all skills that have fixtures/eval.json
./run.sh eval

# Specific skills
./run.sh eval --skill taxonomy,normalize

# Filter by tag
./run.sh eval --tags fast
Installs
2
GitHub Stars
7
First Seen
Mar 17, 2026
eval-skills — grahama1970/agent-skills