Guides the creation of custom, non-interactive agent configurations for specialized Claude Code sessions.
The AI-CLI Modes Authoring skill provides a standardized framework for building 'modes'—preconfigured agent settings for the ai-cli tool. It enables developers to define locked-in system prompts, model selections, and prompt wrappers to create consistent, fire-and-forget workflows. Whether you need a security auditor that never edits code or a debugger that follows a specific methodology, this skill ensures your custom AI personas are structured correctly for both local and global use.
주요 기능
01Support for 'append' or 'replace' logic to toggle between standard tools and custom personas
02Prompt wrapper templates for structuring user input with {{prompt}} placeholders
03Defined frontmatter for model selection and system-prompt-mode orchestration
0424 GitHub stars
05Hierarchical file precedence for project-local and built-in configurations
06Best practices for defining agent constraints over procedural instructions
사용 사례
01Setting up project-specific security auditors that focus on specific vulnerability types
02Automating specialized code reviews within CI/CD pipelines
03Creating 'Investigate-Only' modes that diagnose bugs without modifying the codebase
