The Agent Skills Directory

[DATA_EXPOSURE_AND_EXFILTRATION]: The skill requires all output to be a single, self-contained file and explicitly prohibits network requests, CDNs, and external fonts, effectively preventing data exfiltration during the rendering process.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided content to generate art-directed HTML designs. Ingestion point: user-supplied text. Boundary markers: none. Capability inventory: SendUserFile (render). Sanitization: none. The risk is considered safe as the output is a localized, static document intended for user review.
[SAFE]: JavaScript components for theme switching and scroll animations utilize standard browser APIs like localStorage and IntersectionObserver, following best practices for safe client-side execution.

prettier-html