create-agent-skills

Guide creation of Claude Code skills and slash commands with YAML frontmatter.

49|4|Updated Feb 12, 2026
One-click install
npx skills add https://github.com/gvkhosla/compound-engineering-pi --skill create-agent-skills-gvkhosla
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: create-agent-skills
Source: https://github.com/gvkhosla/compound-engineering-pi/tree/main/plugins/compound-engineering/skills/create-agent-skills
Command: npx skills add https://github.com/gvkhosla/compound-engineering-pi --skill create-agent-skills-gvkhosla

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides expert guidance on creating effective Claude Code skills and slash commands, ensuring your custom tools are discoverable, functional, and adhere to best practices.

Core Features & Use Cases

  • Skill Creation: Learn the structure and syntax for authoring new skills.
  • Command Integration: Understand how custom slash commands are now part of the skill system.
  • Best Practices: Discover how to write effective descriptions, manage dependencies, and structure your skill files for optimal performance.
  • Use Case: You need to build a new /triage-issues command for your team. This Skill will guide you through creating the SKILL.md file, defining its behavior, and ensuring it integrates seamlessly with Claude.

Quick Start

Follow the instructions in this skill to create a new skill file for managing GitHub pull requests.

Frequently Asked Questions about create-agent-skills

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

FAQPage Schema
How do I create a custom slash command for Claude Code?

Claude Code skills use a structured format with YAML frontmatter and markdown content to define custom slash commands. You create a SKILL.md file that specifies the command's behavior, dependencies, and dynamic context injection rules.

What is the difference between a Claude skill directory and a command file?

A Claude skill directory contains a SKILL.md file and supporting resources for complex behaviors, while a command file is a simpler standalone markdown file. You choose a skill directory when you need progressive disclosure or multiple related commands grouped together.

How do I write effective descriptions for Claude skills?

Writing effective descriptions for Claude skills involves clearly defining the task in the YAML frontmatter so the tool is discoverable. You should follow best practices for managing dependencies and structuring your markdown content to ensure optimal performance and seamless integration.

When should I use progressive disclosure in Claude skill development?

You should use progressive disclosure in Claude skill development when your custom tool requires complex, layered instructions. This technique structures your markdown content to provide dynamic context injection, preventing Claude from being overwhelmed while ensuring detailed behavior when needed.

How do I audit a Claude skill for quality assurance?

You audit a Claude skill for quality assurance by running through a provided checklist that validates your YAML frontmatter, markdown content structure, dependency management, and description clarity. This ensures your custom tools are functional and adhere to best practices.

Do I need any dependencies to build a custom tool with Claude skills?

No dependencies are required to build a custom tool with Claude skills. The system guides you through authoring the SKILL.md file from scratch, covering YAML frontmatter syntax and markdown content structure without relying on external libraries or frameworks.