skill-creator

To create an AgentSkill, use a template-driven approach that scaffolds a SKILL.md file with required YAML frontmatter (name and description) and a procedural body, alongside optional resource folders like scripts/ and references/.

Auditing existing agent skills involves checking alignment with specifications, restructuring directories by moving files to references/ or scripts/, and removing stale content to ensure consistent quality and shared structure.

YAML frontmatter in skill documentation enforces required metadata by defining the name and description fields at the top of SKILL.md, ensuring every skill directory adheres to a shared quality standard.

Yes, PyYAML is required as a dependency to parse and enforce the YAML frontmatter metadata when scaffolding new agent skills or auditing existing skill directories.

Move content to references/ or scripts/ during skill refactoring when separating procedural instructions in SKILL.md from supporting deterministic tasks, static documentation, or executable assets.

Yes, the skill-creator workflow supports restructuring stale skill directories by auditing content, removing outdated files, and reorganizing remaining assets into the appropriate references/, scripts/, or assets/ folders.