skills/nirholas/three.ws/auto-rig/Gen Agent Trust Hub

auto-rig

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill defines tools that fetch 3D models and images from user-provided URLs for rigging and generation tasks.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from external sources, creating a potential surface for indirect prompt injection.\n
  • Ingestion points: Data enters via the glb_url, image_url, and prompt parameters in SKILL.md.\n
  • Boundary markers: The instructions lack specific delimiters or warnings to ignore instructions embedded in fetched content.\n
  • Capability inventory: The skill uses tools to perform network operations and generative tasks on the ingested data.\n
  • Sanitization: There are no sanitization or verification steps defined for the external resources.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 04:15 PM
Security Audit — agent-trust-hub — auto-rig