skill-creator

To create a Claude Skill, use a scaffold that enforces YAML frontmatter with name and description, requires a SKILL.md body, and supports optional scripts, references, and assets directories. This structure standardizes skill creation and reduces setup time.

Progressive disclosure in skill packaging is a design pattern that structures resources like scripts and references to reveal complexity gradually. It is needed when building robust skills to prevent overwhelming users while maintaining full functional capabilities.

You can initialize a skill scaffold by running the bundled initialization script with a skill name and target path. This generates the required SKILL.md, frontmatter guidelines, and structured resource directories with example placeholders.

Yes, the skill scaffolding workflow requires PyYAML to parse and enforce the required YAML frontmatter. This dependency ensures the generated skill metadata is correctly formatted for safety and packaging readiness evaluations.

The best way to update existing Claude Skills is applying a reusable scaffold that evaluates YAML frontmatter integrity, checks progressive disclosure, and validates resource directories. This ensures safety and packaging readiness without breaking the existing structure.

Skill packaging readiness evaluations fail when the scaffold detects missing or malformed YAML frontmatter, incomplete SKILL.md bodies, or improperly structured resource directories. Correcting these elements ensures the skill passes safety and packaging checks.