create-agent-skills

Creating custom slash commands for Claude Code involves defining the skill structure, adding frontmatter, and applying progressive disclosure to ensure discoverability. This process ensures the resulting commands are maintainable and function effectively within the agent environment.

Best practices for Claude Code skill creation include using structured frontmatter, applying progressive disclosure to manage complexity, and avoiding common anti-patterns. Adhering to these specifications ensures AI tools remain discoverable, maintainable, and highly effective for agent development.

Auditing prompt engineering workflows involves checking existing skills against official specifications and identifying common anti-patterns. This process evaluates the skill structure, frontmatter, and progressive disclosure to ensure structured and maintainable AI agent development.

Yes, you can add reference files, scripts, or templates to existing Claude skills. This involves updating the skill structure and components to integrate the new references, ensuring the workflow automation tool remains discoverable and functional for agent development.

Progressive disclosure in agent development is needed when managing complex skill structures to prevent overwhelming the AI. It involves structuring frontmatter and references so the agent receives context incrementally, ensuring maintainable and effective workflow automation.