skills/firzus/agent-skills/greyboxing/Gen Agent Trust Hub

greyboxing

Pass

Audited by Gen Agent Trust Hub on Aug 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides detailed, professional-grade instructions for front-end development, specifically focusing on structural layout (greyboxing), motion design, and 3D integration. It incorporates accessibility standards (prefers-reduced-motion) and performance best practices (OffscreenCanvas, LCP optimization).
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external files such as DESIGN.md and PAGES.md. While these are potential ingestion points for untrusted data, the skill includes rigorous review protocols in review.md that prioritize mechanical checklists and visual verification via browser automation, which acts as a safeguard against unexpected behavior.
  • [EXTERNAL_DOWNLOADS]: The documentation references several well-known and industry-standard web libraries including three.js, GSAP, and Lenis. These are used appropriately for implementing 3D effects and complex scroll animations within the generated code.
  • [COMMAND_EXECUTION]: The skill utilizes browser automation tools like agent-browser and playwright-cli to perform visual audits, accessibility checks, and responsive design verification. These tool calls are scoped to testing the UI generated by the agent and do not involve arbitrary shell command execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 16, 2026, 10:47 PM
Security Audit — agent-trust-hub — greyboxing