3d-immersive

Pass

Audited by Gen Agent Trust Hub on May 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves as a development toolbox for React-based 3D animation. The instructions provided under the 'Non-Negotiables' section focus on industry best practices such as accessibility (prefers-reduced-motion), performance optimization (60fps targets, lazy loading), and proper resource management (memory disposal in Three.js).
  • [EXTERNAL_DOWNLOADS]: The skill references several well-known and trusted NPM packages including GSAP, Three.js, React Three Fiber, and Framer Motion. These are standard libraries in the web development ecosystem. All references to external assets (Spline URLs, .glb models) are consistent with the intended purpose of rendering 3D content.
  • [COMMAND_EXECUTION]: The skill provides code snippets for UI components and React hooks. There are no shell commands, system calls, or attempts to execute code within the agent's own execution environment.
  • [DATA_EXFILTRATION]: No patterns associated with data collection, credential harvesting, or network transmission to suspicious domains were found. The use of hooks like useIs3DCapable is strictly for client-side feature detection to provide mobile fallbacks.
  • [PROMPT_INJECTION]: The skill instructions do not contain attempts to override the agent's underlying safety guardrails or system prompts. The 'Non-Negotiables' are specific constraints for the code-generation task.
Audit Metadata
Risk Level
SAFE
Analyzed
May 6, 2026, 01:52 AM
Security Audit — agent-trust-hub — 3d-immersive