Skill Development

Design Claude Code plugin skills with SKILL.md frontmatter and progressive disclosure.

Updated Mar 27, 2026
One-click install
npx skills add https://github.com/quanngynx/GDGO-2026.Servexa-Warranty-AI --skill skill-development-quanngynx
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Skill Development
Source: https://github.com/quanngynx/GDGO-2026.Servexa-Warranty-AI/tree/main/servexa-warranty-ai/.cursor/skills/skill-development
Command: npx skills add https://github.com/quanngynx/GDGO-2026.Servexa-Warranty-AI --skill skill-development-quanngynx

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill helps structure, design, and write effective Claude Code plugin skills so they can be discovered reliably and used correctly by another AI instance.

Core Features & Use Cases

  • Skill structure guidance: Break skill creation into a clear anatomy of required SKILL.md frontmatter plus operational markdown body.
  • Progressive disclosure best practices: Keep SKILL.md lean while pushing detailed patterns, advanced techniques, and large content into references/.
  • Resource planning: Decide when to add scripts/ for deterministic execution, references/ for on-demand context, and assets/ for output files.
  • Quality and validation checklist: Apply trigger phrase design, third-person description rules, and imperative/infinitive writing style to reduce mismatches and improve usability.

Quick Start

Ask the AI to help create a new skill by proposing a directory layout and a draft SKILL.md with third-person YAML frontmatter and specific trigger phrases based on the user’s intended workflow.

Frequently Asked Questions about Skill Development

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

FAQPage Schema
How do I structure a Claude Code plugin skill so it triggers correctly?

Progressive disclosure keeps the main SKILL.md file lean by moving large content and advanced patterns into a references directory. This ensures the AI instance loads only necessary context on demand, maintaining performance while preserving detailed documentation.

What's the best way to organize content when designing Claude Code skills?

Use a scripts directory for deterministic execution tasks, a references directory for on-demand context, and an assets directory for output files. This resource planning separates automated execution from contextual documentation within the Claude Code plugin.

Why does my Claude Code skill description fail to match user requests?

Claude Code skills require YAML frontmatter with name and description fields, a markdown body with imperative instructions, and optional directories including references, scripts, and assets. No external dependencies are needed to start plugin development.

When should I not put all my documentation directly in SKILL.md?

You should not put all documentation in SKILL.md when it contains large content or advanced techniques. Move detailed patterns into a references directory to keep the main file lean and ensure the AI instance can parse operational instructions without loading unnecessary context.