skills/mengto/skills/number-details/Gen Agent Trust Hub

number-details

Pass

Audited by Gen Agent Trust Hub on Jul 25, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The demo template and source metadata reference external resources from well-known and reputable services.
  • Fetches Tailwind CSS from cdn.tailwindcss.com.
  • Loads animation and 3D graphics libraries (GSAP, Three.js) from cdnjs.cloudflare.com.
  • Includes icon assets from code.iconify.design.
  • Downloads web fonts from fonts.googleapis.com and fonts.gstatic.com.
  • [PROMPT_INJECTION]: The skill instructions and demo prompts facilitate the creation of web content based on user-supplied strings, creating an indirect prompt injection surface.
  • Ingestion points: User-provided descriptions, brand names, and subject tags in demo/PROMPT.md.
  • Boundary markers: Absent; there are no clear delimiters used to separate instructions from user-controlled data.
  • Capability inventory: The skill encourages the agent to generate complete HTML documents (index.html) which involves file-writing capabilities.
  • Sanitization: No input validation or sanitization logic is prescribed for the data processed during page generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 25, 2026, 05:38 AM
Security Audit — agent-trust-hub — number-details