structuring-plugins

This skill acts as an architect for developers building extensible modules for the Claude Code ecosystem. It provides comprehensive guidance on adhering to the standardized directory layout required for seamless integration and automatic component discovery. By following these patterns, developers can ensure their plugins maintain consistent structure, portable path references using ${CLAUDE_PLUGIN_ROOT}, and robust manifest configurations.

• Standardizes plugin directory structures including .claude-plugin/plugin.json, commands/, agents/, skills/, and hooks/.

• Manages complex configuration for MCP servers, hook event listeners (PreToolUse, SessionStart, etc.), and custom component path overrides.

• Provides scaffolding workflows for defining sub-agents and modular skills that Claude Code can autonomously activate based on context.

• Ensures naming conventions like kebab-case are strictly followed for reliability and maintainability.

• Automates the integration of helper scripts and utility files within the plugin root.

• Ideal for developers extending Claude Code functionality with custom tools or specialized agentic workflows.

• Useful for setting up plugin manifest fields such as versioning, author metadata, and dependency keywords.

• Supports both simple slash-command extensions and complex event-driven hooks that respond to LLM lifecycle events.

• Constraints: Plugins must adhere to the specified root-level component architecture for automatic discovery to function correctly.

• Inputs: User requirements for a new feature, directory reorganization, or manifest update. Output: Scaffolding instructions, file system patterns, or validation of existing plugin structure.