blender-director
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, unauthorized data exfiltration, or persistence mechanisms were found. The skill operates within a professional production framework for 3D assets.
- [COMMAND_EXECUTION]: The skill utilizes a dedicated MCP server to execute Blender Python (bpy) commands. This is used for legitimate automation tasks such as collection management, mesh cleanup, and file export, following best practices for technical art pipelines.
- [EXTERNAL_DOWNLOADS]: References the official blender-mcp repository on GitHub (github.com/ahujasid/blender-mcp) as a trusted source for setup and tool capabilities.
- [PROMPT_INJECTION]: The skill processes untrusted user data (asset descriptions and reference images) which represents an indirect prompt injection surface. Ingestion points: User asset requests and reference images (SKILL.md). Boundary markers: Absent; instructions do not explicitly mandate delimiters for user-provided data. Capability inventory: Python code execution and file writes via the Blender MCP server (references/mcp-tools.md). Sanitization: Absent; however, all operations are constrained to the Blender modeling environment.
Audit Metadata