webgl-particle-systems

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides legitimate technical documentation and code samples for Three.js WebGL development, following standard industry practices for GPU-driven animations.
  • [EXTERNAL_DOWNLOADS]: The skill references the Three.js library via the official npm registry (npm i three) and the JSDelivr CDN. These are well-known and trusted sources for web development dependencies.
  • [INDIRECT_PROMPT_INJECTION]: The skill includes a sampleText function that processes strings to generate particle layouts. While this involves ingesting external data, the function only performs pixel sampling on a 2D canvas for visual representation and does not interpolate the text into agent instructions or executable logic.
  • Ingestion points: sampleText(text, count, width, height) function in SKILL.md.
  • Boundary markers: None (typical for visual data processing).
  • Capability inventory: Limited to canvas rendering and WebGL buffer updates.
  • Sanitization: Not applicable as the input is strictly used for visual pixel data.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 05:48 PM
Security Audit — agent-trust-hub — webgl-particle-systems