Claude Skills
CollectionsCompareWorkflowsNominate
Sign inSign up
© 2026 Curated Agent Skills·Learn more about Agent Skills
Back to repository

claude-skills

verified

Guide for creating Claude Agent Skills. Use when designing Skills, writing SKILL.md files, structuring Skill directories, or understanding progressive disclosure, metadata, and best practices for reusable domain-specific capabilities.

View on GitHub

Marketplace

maledorak-marketplace

maledorak/maledorak-marketplace

Plugin

claude-toolkit

Repository

maledorak/maledorak-marketplace
1stars

plugins/claude-toolkit/skills/claude-skills/SKILL.md

Last Verified

February 1, 2026

Install Skill

Select agents to install to:

Scope:
npx add-skill https://github.com/maledorak/maledorak-marketplace/blob/main/plugins/claude-toolkit/skills/claude-skills/SKILL.md -a claude-code --skill claude-skills

Installation paths:

Claude
.claude/skills/claude-skills/
Powered by add-skill CLI

Instructions

# Claude Agent Skills Creation Guide

Create modular, reusable Agent Skills that extend Claude's functionality with domain-specific expertise.

## What are Agent Skills?

**Agent Skills** are filesystem-based capabilities that provide Claude with domain-specific expertise: workflows, context, and best practices. Skills load on-demand and eliminate repeated guidance across conversations.

**Key benefits:**
- **Specialize Claude**: Tailor capabilities for specific tasks
- **Reduce repetition**: Create once, use automatically
- **Compose capabilities**: Combine Skills for complex workflows
- **Progressive disclosure**: Load content only as needed

## Quick Start

### Minimal Skill Structure

```
my-skill/
└── SKILL.md
```

**SKILL.md format:**
```yaml
---
name: my-skill
description: Brief description of when to use this Skill. Include keywords that trigger relevance.
---

# My Skill Name

Quick start content, workflows, and examples.

## Common Task

Step-by-step instructions...
```

### With Reference Files

```
my-skill/
├── SKILL.md           # Main entry point
└── reference/
    ├── guide.md       # Detailed documentation
    └── examples.md    # Comprehensive examples
```

## Core Principles

### 1. Progressive Disclosure (3 Levels)

**Level 1: Metadata** (always loaded)
- YAML frontmatter with `name` and `description`
- Loaded at startup, included in system prompt
- Lightweight - no context penalty
- **Use for:** Skill discovery ("when to use this")

**Level 2: Instructions** (loaded when triggered)
- Main body of SKILL.md
- Loaded when Skill becomes relevant
- **Use for:** Workflows, best practices, guidance

**Level 3: Resources** (loaded as needed)
- Additional markdown files, scripts, templates
- Loaded only when referenced
- **Use for:** Detailed docs, code utilities, reference materials

### 2. Concise is Key

**Challenge every token:**
- Does Claude really need this explanation?
- Can I assume Claude knows this?
- Does this justify its context cost?

**Goo

Validation Details

Front Matter
Required Fields
Valid Name Format
Valid Description
Has Sections
Allowed Tools
Instruction Length:
11416 chars