ai-elements

To create an AI Elements Skill Unit, you must add a SKILL.md file containing YAML frontmatter with a name and description, alongside any optional scripts or assets, to generate structured metadata for discovery and safety checks.

The required SKILL.md frontmatter metadata includes a name and description to enable search indexing, while optional fields document dependencies on scripts, references, and assets to map the unit's tooling structure.

You document optional scripts and references by placing them in their respective directories and running the metadata generator, which analyzes these resources to produce a YAML manifest detailing the component's structure and dependencies.

Yes, adding a new Skill Unit requires a toxicity check for safety validation, ensuring the generated metadata and component content meet safety standards before being indexed within the AI Elements project.

The best way to structure component tooling is to organize resources into separate directories for scripts, references, and assets, allowing the metadata generator to accurately analyze dependencies and produce the YAML manifest.

Yes, you can use nanoid and lucide-react as dependencies when building AI Elements components, declaring them in the SKILL.md frontmatter so the metadata generator can map them for tooling discovery and structural analysis.