skill-creator

To create modular workflows, you build self-contained Skills using a compact frontmatter-based trigger and body-guided instructions. This structure packages domain knowledge and tool integrations into reusable artifacts with bundled resources like scripts and references.

The best way to structure Codex skills is using a compact frontmatter trigger, a body-guided instruction section, and optional bundled resources. This design balances context window usage and operational reliability for rule-based knowledge artifacts.

You scaffold a new skill by running the initialization script with a chosen name and output directory. This generates a modular structure with optional bundled resources like scripts and references, establishing a clear lifecycle from init to iteration.

Yes, PyYAML is a required dependency for skill creation. It processes the frontmatter-based templates and guidelines that define the compact triggers and body-guided instructions within the modular skill artifacts.

Yes, you can update existing modular skills through a clear lifecycle process of editing, validating, and iterating. This maintains consistency and reusability across projects while preserving bundled resources and tool integrations.

You should bundle resources like scripts and references when your workflow requires repeatable, rule-based knowledge artifacts. Bundling these assets alongside the frontmatter and body instructions ensures the skill remains self-contained and operational.