blender-workflow-routing
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process externally supplied or untrusted scene content. It mitigates this risk by explicitly instructing the agent to route such content through a safety skill (
blender-content-safety-and-privacy) before opening or importing it. - Ingestion points: Externally supplied scene files (SKILL.md).
- Boundary markers: The skill instructs the agent to use
blender-content-safety-and-privacyas a boundary. - Capability inventory: The skill routes tasks to
blender-python-executionfor script running. - Sanitization: Instructions require using
blender-sanitized-support-bundleto clean logs and manifests before sharing. - [DATA_EXPOSURE_AND_EXFILTRATION]: The skill enforces privacy best practices by requiring that support bundles, logs, and screenshots be sanitized and that unsanitized artifacts remain local.
- [COMMAND_EXECUTION]: The skill routes to
blender-python-executionfor agent-controlled Blender sessions. This is a standard functional requirement for Blender automation and is governed by internal routing rather than arbitrary external command strings.
Audit Metadata