plugin-creator

# Claude Code Plugin Creator

## Overview

This skill provides comprehensive guidance for creating Claude Code plugins following the official Anthropic format (as of December 2025).

## Plugin Architecture

A Claude Code plugin can contain any combination of:
- **Commands**: Custom slash commands (`/mycommand`)
- **Agents**: Specialized AI subagents for specific tasks
- **Hooks**: Pre/post tool execution behaviors
- **MCP Servers**: Model Context Protocol integrations
- **Skills**: Domain-specific knowledge packages

## Directory Structure

```
my-plugin/
├── .claude-plugin/
│   └── plugin.json          # REQUIRED - Plugin manifest
├── commands/                 # Optional - Slash commands
│   └── my-command.md
├── agents/                   # Optional - Subagents
│   └── my-agent.md
├── hooks/                    # Optional - Hook definitions
│   └── hooks.json
├── skills/                   # Optional - Bundled skills
│   └── my-skill/
│       └── SKILL.md
├── mcp/                      # Optional - MCP server configs
└── README.md
```

## Plugin Manifest (plugin.json)

The `.claude-plugin/plugin.json` file is **required**. Here's the complete schema:

```json
{
  "$schema": "https://anthropic.com/claude-code/plugin.schema.json",
  "name": "my-plugin",
  "version": "1.0.0",
  "description": "Clear description of what this plugin does",
  "author": {
    "name": "Your Name",
    "email": "you@example.com"
  },
  "license": "MIT",
  "commands": [
    {
      "name": "mycommand",
      "description": "What this command does",
      "source": "./commands/my-command.md"
    }
  ],
  "agents": [
    {
      "name": "my-agent",
      "description": "What this agent specializes in",
      "source": "./agents/my-agent.md"
    }
  ],
  "hooks": {
    "source": "./hooks/hooks.json"
  },
  "skills": [
    "./skills/my-skill"
  ],
  "mcp_servers": [
    {
      "name": "my-mcp",
      "command": "npx",
      "args": ["my-mcp-server"]
    }
  ]
}
```

## Creating Commands

Commands