web-design

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [SAFE]: The skill focuses on architectural and design principles for building websites, requiring the agent to validate the brand state and copy before initiating design. It promotes best practices such as accessibility, fluid typography (using rem and clamp), and performance-conscious motion design.- [EXTERNAL_DOWNLOADS]: The skill references industry-standard libraries and tools for enrichment.
  • External Libraries: Mentions using GSAP (GreenSock Animation Platform) and three.js/WebGL for animation and interactive depth.
  • Design Accelerators: Integrates with Relume MCP and Mobbin/Refero MCP for sitemap generation, wireframing, and pattern intake.
  • Asset Sources: Recommends pulling official brand marks from simple-icons or source site CDNs.- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests data from external websites and local files to generate code.
  • Ingestion points: Processes external reference URLs provided by the user and local files in the .agents/ directory (design.md, brand.md, product-marketing.md).
  • Boundary markers: No specific delimiters or 'ignore' instructions for instructions embedded in external content are documented.
  • Capability inventory: The skill generates and writes production-grade HTML, CSS, and JavaScript to the project filesystem.
  • Sanitization: No explicit sanitization or filtering logic for processed external content is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 12:31 PM