create-skill

Guide users in authoring custom Agent Skills for Cursor with metadata and structure.

Updated Feb 20, 2026
One-click install
npx skills add https://github.com/garethdaine/agent --skill create-skill-garethdaine
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: create-skill
Source: https://github.com/garethdaine/agent/tree/main/.cursor/skills-cursor/create-skill
Command: npx skills add https://github.com/garethdaine/agent --skill create-skill-garethdaine

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill guides users through the process of creating well-structured and effective Agent Skills for Cursor, ensuring they are discoverable and performant.

Core Features & Use Cases

  • Skill Structure Guidance: Provides best practices for directory layout, SKILL.md format, and metadata.
  • Best Practice Enforcement: Details how to write effective descriptions, manage context window, and use progressive disclosure.
  • Use Case: A user wants to create a new skill to automatically generate commit messages based on git diffs. This skill will guide them through defining the skill's name, description, and structure, and suggest best practices for the SKILL.md content.

Quick Start

Guide me through creating a new skill for analyzing code quality.

Frequently Asked Questions about create-skill

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

FAQPage Schema
How do I create custom agent skills for Cursor?

To create agent skills for Cursor, you define a directory layout, write a SKILL.md file, and add metadata. This ensures your custom skill is discoverable and performs effectively within the agent workflow.

What is the best way to structure an agent skill for optimal context window usage?

The best way to structure an agent skill is using progressive disclosure in your SKILL.md file. This approach manages the context window by providing essential metadata upfront and detailed instructions only when needed.

How does progressive disclosure work when writing agent skill documentation?

Progressive disclosure works by layering skill documentation, starting with concise descriptions and metadata. It reveals deeper structural details and workflows only when the agent requires them, preventing context window overload.

Do I need to follow specific naming conventions and metadata requirements for Cursor agent skills?

Yes, you must follow specific naming conventions and metadata requirements for Cursor agent skills. Adhering to these standards ensures your skill is properly indexed, discoverable, and clearly understood by the agent.

What should I include in a SKILL.md file to ensure my custom skill is performant?

A SKILL.md file should include clear descriptions, proper metadata, and progressive disclosure principles. Organizing content this way optimizes agent understanding and ensures your custom skill performs reliably.