This skill provides comprehensive, expert guidance for building custom extensions within the Claude Code ecosystem. It bridges the gap between simple slash commands and complex skill directories, offering best practices for YAML frontmatter configuration, invocation control, and dynamic context injection. By following this skill's structured methodology, developers can create more reliable, discoverable, and efficient AI workflows that automate repetitive tasks like PR summaries, deployments, and code audits while adhering to official performance standards.
主要功能
01Guidance on progressive disclosure for complex skill structures
02Dynamic context injection using shell commands and placeholders
03Standardized markdown formatting and audit checklists
04YAML frontmatter configuration for precise invocation control
0535 GitHub stars
06Subagent forking for isolated task execution