Skills Engineering

Author, evaluate, and iteratively improve Agent Skills (SKILL.md protocol) using a rigorous TDD lifecycle. Handles spec declaration, BDD behavioral contract authoring, script TDD (red-green per acceptance criterion), SKILL.md authoring, adversarial boundary testing, full eval with subagent-driven grading, improvement loops, and optional refactoring. Use when creating a new skill, evaluating an existing skill, improving a skill from eval feedback, or engineering the SKILL.md structure.

cristoslc Updated 2 repo stars

File contents

cristoslc/skills-engineering-skill commit aeabf7d272

Frequently asked questions

npx get-skill add cristoslc/skills-engineering