KillerSkills
Sign in
Documentation

SKILL.md Format Reference

The SKILL.md format is a standardized specification for defining AI agent skills. Each skill file consists of YAML frontmatter followed by markdown content.

Complete Example

---
name: React Component Generator
description: Generate production-ready React components with TypeScript, props interface, and documentation
version: "1.0.0"
author: KillerSkills Team
runtimes: [claude-code, cursor, copilot]
category: development
tags: [react, typescript, components, frontend]
tools: [typescript, prettier]
triggers: [create component, generate react component]
inputs:
  - name: componentName
    type: string
    description: Name of the component to generate
    required: true
  - name: componentType
    type: string
    description: Type of component (functional, class, hook)
    required: false
outputs:
  - name: componentFile
    type: string
    description: Generated component TypeScript file
  - name: testFile
    type: string
    description: Generated test file for the component
---

# React Component Generator

This skill generates production-ready React components with TypeScript type definitions,
proper props interfaces, and comprehensive documentation.

## Usage

When the user requests a new React component, follow these steps:

1. **Analyze Requirements**: Understand the component's purpose, required props, and functionality
2. **Generate Interface**: Create a TypeScript interface for component props
3. **Create Component**: Generate a functional component with proper typing
4. **Add Documentation**: Include JSDoc comments explaining props and usage
5. **Generate Tests**: Create basic test file with example test cases

## Example

For a Button component:

```typescript
interface ButtonProps {
  label: string;
  onClick: () => void;
  variant?: 'primary' | 'secondary';
  disabled?: boolean;
}

/**
 * Button component with multiple variants
 * @param props - Button props
 */
export const Button: React.FC<ButtonProps> = ({
  label,
  onClick,
  variant = 'primary',
  disabled = false,
}) => {
  return (
    <button
      onClick={onClick}
      disabled={disabled}
      className={`btn btn-${variant}`}
    >
      {label}
    </button>
  );
};
```

## Best Practices

- Always use TypeScript for type safety
- Include default values for optional props
- Add JSDoc comments for documentation
- Follow React naming conventions
- Generate corresponding test files

Frontmatter Fields Reference

The YAML frontmatter section contains metadata about the skill. All fields are optional unless marked as required.

FieldTypeRequiredDescription
namestringYesShort, descriptive name for the skill. Must not be empty.
descriptionstringNoBrief description of what the skill does. Maximum 500 characters.
versionstringNoSemantic version number (e.g., "1.0.0"). Should follow semver conventions.
authorstringNoName or username of the skill author.
runtimesstring[]NoList of compatible AI runtimes. Valid values: claude-code, cursor, copilot, codex-cli, chatgpt, gemini-cli, windsurf, aider, continue, cline, other.
categorystringNoPrimary category for the skill. Valid values: development, devops, writing, data, design, testing, security, documentation, automation, other.
tagsstring[]NoList of tags for discovery and search. Maximum 10 tags, each tag maximum 30 characters.
toolsstring[]NoList of external tools or dependencies the skill uses (e.g., "typescript", "prettier", "docker").
triggersstring[]NoConditions or phrases that should activate this skill (e.g., "create component", "write tests").
inputsobject[]NoArray of input parameter definitions. Each input has: name (string), type (string), description (string), required (boolean).
outputsobject[]NoArray of output definitions. Each output has: name (string), type (string), description (string).

Body Content

The markdown body follows the YAML frontmatter (after the closing ---). This section contains the detailed instructions that guide the AI agent in performing the task.

Requirements

  • Must contain at least 50 characters of meaningful content
  • Should include clear, step-by-step instructions
  • Can use standard markdown formatting (headings, lists, code blocks, etc.)
  • Should provide examples demonstrating the skill in action
  • May include usage guidelines, best practices, and edge cases

Structure Guidelines

While not required, effective skill bodies typically include:

  • Overview — brief introduction to what the skill accomplishes
  • Usage Instructions — step-by-step guide for the AI agent
  • Examples — concrete code examples or scenarios
  • Best Practices — recommendations for optimal results
  • Edge Cases — how to handle unusual or complex situations

Validation Rules

The platform validates all skills before they can be published. Validation produces two types of feedback:

Errors (Must Fix)

  • name field is required and must not be empty
  • Body content must exist and contain meaningful text
  • If version is provided, it should follow semantic versioning (e.g., "1.0.0", "2.1.3")
  • description must not exceed 500 characters
  • Maximum 10 tags, each tag maximum 30 characters
  • runtimes must only contain valid runtime identifiers
  • category must be one of the allowed category values

Warnings (Best Practice)

  • Body content should be at least 50 characters for meaningful instructions
  • Consider adding a description for better discoverability
  • Consider specifying compatible runtimes
  • Consider adding tags to improve search relevance
  • Consider specifying a category for organization

Next Steps