three-postprocessing-stack

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions include standard package management commands (npm install) to install well-known and version-pinned dependencies (three, postprocessing, @react-three/postprocessing). These are legitimate development tools.
  • [DYNAMIC_EXECUTION]: The implementation uses ResizeObserver and GPU timer queries (EXT_disjoint_timer_query_webgl2) to monitor frame times and adjust rendering quality tiers. These are standard WebGL performance optimization techniques and do not involve untrusted code execution.
  • [DATA_EXFILTRATION]: No network calls or data transmission patterns were identified. The code focuses exclusively on the local 3D rendering pipeline and performance monitoring.
  • [PERSISTENCE]: The skill emphasizes a comprehensive dispose() pattern to ensure that GPU resources (render targets, textures, and shader programs) are explicitly released, preventing memory leaks and unintended resource persistence in the browser environment.
  • [SAFE]: The content is purely educational and technical, providing clear guidance on avoiding common rendering pitfalls like double tone-mapping or redundant anti-aliasing passes.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 05:48 PM
Security Audit — agent-trust-hub — three-postprocessing-stack