pretext

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides templates and instructions for creating browser-based typography demos. No malicious code or instructions were found.
  • [EXTERNAL_DOWNLOADS]: Dependencies like the pretext library, GSAP, and lil-gui are fetched from well-known CDNs (esm.sh and jsdelivr). These are reputable sources for development resources.
  • [DATA_EXFILTRATION]: No sensitive local data access or unauthorized network exfiltration was detected.
  • [PROMPT_INJECTION]: No behavioral override patterns or safety bypass attempts were observed in the skill's instructions.
  • [SAFE]: Evaluation of indirect prompt injection surfaces (Category 8): Ingestion point: Text corpus processed in templates. Boundary markers: None. Capabilities: Generates HTML/JS for text rendering. Sanitization: HTML entities are escaped using an escapeHTML function in the demo templates to prevent XSS.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 01:37 PM
Security Audit — agent-trust-hub — pretext