Claude Extensibility

소개

Provides a comprehensive framework for extending Claude Code's capabilities through the creation and management of specialized agents, skills, and output styles. It enforces critical development standards like the 500-line rule and progressive disclosure to ensure high-performance, token-efficient interactions. By streamlining tool access configuration, model selection, and prompt optimization, this skill allows users to tailor Claude's behavior to specific project requirements while maintaining peak efficiency and reliability.

주요 기능

• Support for resumable agents with unique task IDs and transcript tracking
• Agent and skill lifecycle management including creation, optimization, and testing
• YAML frontmatter configuration for model selection and tool permission modes
• Implementation of WHAT + WHEN description patterns for reliable skill activation
• Customizable output styles to modify the main agent's CLI behavior
• 5 GitHub stars

사용 사례

• Optimizing extension performance by following Anthropic's efficiency guidelines and prompt structures
• Creating specialized sub-agents for complex research, planning, or code review tasks
• Building domain-specific skill modules to provide Claude with project-specific knowledge