agent-skills-writing

Standardize AI agent skill writing by applying required frontmatter, folder structure conventions, and reference extraction patterns to ensure automated discovery and maintainability across user-facing and internal guidelines.

Organize reference files by extracting them into dedicated directories using hub and reference workflows, enabling consolidated skills to maintain clear navigation hubs without bloating the core SKILL.md file.

The skill-creator workflow evaluation checklist specifies validation requirements for frontmatter, folder structure, and reference extraction, ensuring new skills meet repository standards before registration and automated discovery.

Use consolidated patterns when a skill grows large and requires reference-file strategies, while standalone patterns suit simpler skills without optional resources like scripts, references, or assets directories.

Internal guidelines reside under .github/skills/ while user-facing skills occupy skills/, both requiring standard frontmatter and navigation hub conventions to enable automated validation and consistent discoverability.

Consolidation requires careful reference-file extraction and hub management to avoid navigation complexity, demanding strict adherence to frontmatter standards and folder structure conventions to maintain automated discovery capabilities.