hard-surface
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill instructs the agent to analyze user-provided photos or concepts to extract parameters like color palettes, camera angles, and features, which creates an indirect prompt injection surface.\n- Ingestion points: User-provided images and concepts (references/asset-pipeline.md).\n- Boundary markers: None specified to differentiate between user data and instructions.\n- Capability inventory: The agent can create primitives, apply modifiers (booleans, bevels), and query polycounts via MCP (SKILL.md).\n- Sanitization: No evidence of input validation or sanitization for extracted reference data.
Audit Metadata