What is the Iron Law in skill authoring?

The Iron Law states that no skill should be written without first documenting a failing test case where Claude fails to perform the task correctly without the intervention.

What is Scribe validation?

Scribe validation is a quality gate that checks for 'slop' scores, ensures all file paths work, and maintains a specific bullet-to-prose ratio for maximum readability.

What are anti-rationalization patterns?

These are specific instruction sets and 'red flag' lists that prevent the AI from making excuses to skip complex requirements or claiming a task is too simple for the full process.

How does this help with token management?

It utilizes progressive disclosure and modular architecture to keep individual skill files under 500 lines, ensuring the model only consumes tokens for the specific logic required.

Can I use this to evaluate my current skills?

This skill is specifically for authoring and creation. For evaluating the performance of existing skills, it is recommended to use the 'skills-eval' plugin.

Skill Authoring & TDD Framework

Name: Skill Authoring & TDD Framework
Author: Ven0m0

byVen0m0

0•

学習とドキュメンテーション

Standardizes the creation of high-performance Claude Code skills through empirical test-driven development and persuasion principles.

This skill provides a rigorous framework for developing, validating, and deploying specialized Claude Code capabilities using a mandatory Test-Driven Development (TDD) methodology. It prevents common AI failure modes by requiring documented evidence of failure (RED phase) before implementation (GREEN phase) and includes advanced anti-rationalization techniques to ensure model compliance. Ideal for developers building complex modular skills, it ensures every behavioral intervention is measurable, token-efficient, and bulletproofed against model shortcuts.

主な機能

01Mandatory TDD methodology (RED-GREEN-REFACTOR) for skill development

02Progressive disclosure architecture for efficient token usage

03Automated validation scripts for complexity and token estimation

040 GitHub stars

05Semantic search optimization formula for skill discovery

06Advanced anti-rationalization patterns to prevent model bypass

ユースケース

01Creating high-compliance skills for specialized domain-specific tasks

02Standardizing a skill-writing workflow across a development team

03Debugging and hardening existing skills that Claude frequently ignores

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ven0m0/claude-config skill-authoring

For use in Claude.ai and ChatGPT

Download Skill