skill-creator

You create a reusable agent skill by structuring a SKILL.md file with validated YAML frontmatter, kebab-case naming, and detected dependencies. This packaging process ensures the skill is reliably triggered and executed by other agents.

Structuring an AI agent skill for reuse involves organizing instructions, optional scripts, references, and assets into a cohesive package. Proper structure with validated frontmatter allows another agent to trigger and use the capability reliably.

Yes, you need the pyyaml dependency installed to parse and validate the required YAML frontmatter in SKILL.md files. This dependency allows the skill creation process to automatically verify metadata and detect components from sibling directories.

Use scripts when your agent skill requires automated execution logic, and use references when providing contextual documentation. The skill creation process helps decide the correct mix of instructions, scripts, references, and assets for your specific workflow.

Yes, you can update an existing agent skill package without breaking triggers by re-validating the SKILL.md metadata. The skill update workflow checks frontmatter, naming conventions, and component detection to maintain trigger compatibility.

Agent skill validation fails when the SKILL.md file does not adhere to kebab-case naming requirements. You must rename your skill package using lowercase letters and hyphens to pass the structural validation checks.