scenario-blender-expert
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data such as user-supplied artistic briefs and 3D mesh files, which serves as a potential ingestion point for malicious instructions. Ingestion points: Artistic briefs and 3D models (SKILL.md, bx_audit.py). Boundary markers: None identified. Capability inventory: Full bpy script execution, file system writes for renders/audits, and Blender process execution (bx_review.py, SKILL.md). Sanitization: No explicit sanitization or validation of input metadata or model content is implemented.
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to run Blender in headless mode (e.g.,
blender -b ... --python-exit-code 1). This is a necessary capability for the skill's purpose but represents a controlled execution surface. - [DYNAMIC_EXECUTION]: The skill's primary function is the generation and execution of Python scripts using the Blender
bpyAPI. This involves runtime script assembly and execution, which is an intended feature of the professional 3D workflow provided.
Audit Metadata