The Agent Skills Directory

[OBFUSCATION]: The file demo/index.html contains a large Base64-encoded string (encodedHtml). This string decodes to a full HTML document used for the design demonstration, effectively hiding the source code from static plain-text analysis.
[DYNAMIC_EXECUTION]: The JavaScript within demo/index.html decodes the encodedHtml Base64 blob and dynamically injects the resulting content into a sandboxed iframe's srcdoc attribute. This allows for the execution of hidden HTML and scripts at runtime.
[INDIRECT_PROMPT_INJECTION]: The skill represents a vulnerability surface by ingesting user-provided search tags to query an external image library (aura.build/assets).
Ingestion points: The search query parameter q in the aura.build URL (SKILL.md).
Boundary markers: No explicit instructions are provided to the agent to ignore potentially malicious content returned in search metadata or filenames.
Capability inventory: The skill has the capability to return external image URLs to the user or a downstream application.
Sanitization: There is no evidence of sanitization or validation of the content returned from the external search before it is interpolated into the agent's output.
[EXTERNAL_DOWNLOADS]: The skill's demo component fetches runtime dependencies from several well-known third-party services including Cloudflare (GSAP), Iconify, and Google Fonts. These are used to support the visual and interactive elements of the provided design template.

aura-asset-images