The Agent Skills Directory

[METADATA_POISONING]: The skill's frontmatter includes a self-assessment field risk: safe. This is an authoritative claim about its own safety that was evaluated independently and found to be accurate in this instance.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided descriptions of icon concepts to search for or generate icons using an external tool (app.iconsax.io/ai).
Ingestion points: User-defined concepts for icons in the Search or Generate step.
Boundary markers: Absent from the instructions.
Capability inventory: The skill involves the selection and integration of SVGs or web components into the application's source code.
Sanitization: Not specified in the instructions.

iconsax-library