blender-expert

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The Blender skill’s required runtime workflow ingests user/attacker-authored free text via construct_object(description=...) and agentic_blender_workflow(workflow_prompt=...), which are natural-language prompts used by the LLM to drive Blender actions.