Skills
Navigate
skills/agentpmt/agent-skills/blender-3d-modeling/Gen Agent Trust Hub

blender-3d-modeling

Warn

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: MEDIUMDYNAMIC_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The run_script action allows the execution of arbitrary Python scripts within a remote Blender environment, establishing a dynamic execution surface for agent-generated code.\n- [REMOTE_CODE_EXECUTION]: The skill enables remote logic execution on the vendor's infrastructure through the run_script action.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted 3D model files from external URLs, which serves as a potential vector for indirect prompt injection.\n
  • Ingestion points: file_url and file_id parameters in rendering and processing actions like render_turntable and run_script.\n
  • Boundary markers: None identified for model data ingestion.\n
  • Capability inventory: Remote Python execution (run_script), file conversion, and automated rendering operations.\n
  • Sanitization: No validation or sanitization is mentioned for the internal content of the ingested 3D models.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 11, 2026, 11:38 PM