skill-creator

Package reusable agent skills by creating a SKILL.md-based directory with valid YAML frontmatter, concise operational guidance, and bundled scripts or references for deterministic execution. This structure ensures the assistant triggers consistent, domain-specific behavior without rewriting instructions each time.

YAML frontmatter is metadata at the top of a SKILL.md file containing required name and description fields. It enables skill discoverability and activation, allowing the agent to identify and trigger the correct specialized workflow when matching queries occur.

Bundle scripts and references into an agent skill directory by organizing them alongside the SKILL.md file in a structured layout. This resource planning enables deterministic execution and richer context, supporting domains like document processing, code automation, and API integration.

Yes, you can update existing skills by modifying the operational guidance in the SKILL.md body while maintaining valid YAML frontmatter. Preserving the name and trigger phrases ensures current workflow automations continue activating correctly.

Yes, skill-creator requires the pyyaml Python dependency to parse and validate YAML frontmatter in SKILL.md files. This enables proper skill structure packaging and deterministic execution across document processing and code automation domains.

Avoid using a SKILL.md-based directory for one-off tasks that lack repeatable workflows or domain-specific triggers. Skills are designed for reusable automation patterns like data extraction or API integration, not single-use prompt interactions.