fantasy-character-visual-dna

Pass

Audited by Gen Agent Trust Hub on Oct 3, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface as it processes user-provided character descriptions and reference images. * Ingestion points: User character ideas and reference images as defined in SKILL.md. * Boundary markers: Delimiters and explicit warnings for untrusted data are not specified in the instructions. * Capability inventory: The skill invokes internal image generation tools and produces text prompts; it lacks file write or general network capabilities. * Sanitization: No explicit input sanitization or filtering logic is included.
  • [EXTERNAL_DOWNLOADS]: The documentation includes standard installation instructions utilizing Git to download the skill repository from GitHub. Evidence: Installation commands in README.md target the author's public repository on a well-known hosting service.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 3, 2026, 12:11 AM