ck:skill-creator

To create a Claude skill with a consistent folder structure, use a scaffold to auto-generate a SKILL.md template alongside bundled resource directories like scripts, references, and assets for deterministic task execution.

Progressive disclosure in skill creation is a metadata-driven activation mechanism that enforces structured information layering, ensuring consistent discovery and benchmarking readiness across generated skills.

Bootstrap a new skill for Skillmark benchmarking by executing an initialization workflow that generates a production-grade skeleton in your target directory, which you then tailor to meet benchmark criteria.

Yes, scaffolding Claude skills requires Python, specifically the python-dotenv dependency, to manage environment variables and support the automation of deterministic tasks across the generated resource folders.

Yes, you can use this scaffolding to generate starter templates, as it creates bundled resource folders containing example templates for scripts, references, and assets during the skill initialization workflow.

The best way to structure a SKILL.md file for consistent discovery is using a frontmatter template with metadata-driven activation, which enforces progressive disclosure and ensures your skill meets benchmarking standards.