skill-creator

Enforce test-driven development for creating and editing Claude Code skills.

Updated Nov 7, 2025
One-click install
npx skills add https://github.com/m-brady/claude-plugins --skill skill-creator-m-brady
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/m-brady/claude-plugins/tree/main/plugins/skill-creator-plugin/skills/skill-creator
Command: npx skills add https://github.com/m-brady/claude-plugins --skill skill-creator-m-brady

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

Creating effective AI skills can be challenging, often leading to vague instructions or missed edge cases. This Skill enforces a rigorous Test-Driven Development (TDD) approach, ensuring your skills are precise, reliable, and address real user needs, saving you debugging time.

Core Features & Use Cases

  • TDD for Skills: Guides you through the RED-GREEN-REFACTOR cycle to develop and refine skills, ensuring they perform exactly as intended.
  • Baseline Testing: Emphasizes testing Claude without the skill first to identify exact failure points, so you know precisely what to teach.
  • Anti-Patterns & Loopholes: Helps identify and close rationalizations or shortcuts Claude might take, ensuring disciplined and consistent behavior from your AI.
  • Use Case: You want to create a new skill for code review. This Skill will guide you to first observe Claude's current code review failures, then write a minimal skill to fix those, and finally refine it to prevent new loopholes, resulting in a highly effective and reliable code review skill.

Quick Start

I want to create a new skill for analyzing log files. Guide me through the TDD process for skill creation.

Frequently Asked Questions about skill-creator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I use test-driven development to create AI skills?

Test-driven development for skills means writing tests first to define desired behavior, then building the skill to pass those tests. This Skill guides you through RED-GREEN-REFACTOR cycles: observe Claude's baseline failures without the skill, write minimal skill instructions to fix them, and refine to prevent edge cases and loopholes.

What's the best way to validate skill inputs like name format and description triggers?

Input validation for skills involves checking constraints such as name format standards, length limits, and description requirements before processing. This Skill includes built-in validation logic to enforce these constraints during skill creation and editing.

How do I structure a skill with YAML frontmatter and directory scaffolding?

Skill structure combines YAML frontmatter metadata with organized directories for scripts, references, and assets. This Skill automates directory scaffolding and SKILL.md generation, creating a consistent foundation for your skill project.

Can I test skill behavior before and after making changes?

Yes. This Skill integrates a simple test harness that runs baseline tests before changes and validates behavior after changes, ensuring your skill modifications work correctly and don't introduce regressions.

Why should I test Claude without a skill first when developing one?

Testing Claude's baseline behavior without the skill reveals exact failure points and gaps, giving you precise targets to address. This prevents over-engineering and ensures your skill solves real, observed problems rather than assumed ones.

How do I identify and close loopholes in skill instructions?

Loopholes occur when Claude finds shortcuts or rationalizations that bypass intended behavior. This Skill's anti-pattern guidance helps you spot these gaps through iterative testing and refine instructions to enforce disciplined, consistent AI behavior.

Related Skills