scenario-3d

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied prompts and reference images for 3D generation, which constitutes an indirect prompt injection attack surface. 1. Ingestion points: User prompts and image assets used in tools like recommend and model_run. 2. Boundary markers: No explicit delimiters or boundary markers are instructed for use. 3. Capability inventory: Use of model_run for generation and curl for downloading assets. 4. Sanitization: No input validation or sanitization steps are mentioned.
  • [EXTERNAL_DOWNLOADS]: The instructions describe downloading generated 3D assets (e.g., GLB, FBX files) from remote URLs using curl -L.
  • [COMMAND_EXECUTION]: The documentation mentions using npx skills add for installing additional skill components and curl for downloading asset files.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 07:46 AM
Security Audit — agent-trust-hub — scenario-3d