thatopen-core-architecture
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill serves as a technical reference for the ThatOpen component system. It contains legitimate architectural guidance, API signatures, and code examples for 3D visualization.\n- [INDIRECT_PROMPT_INJECTION]: The skill involves loading and processing Industry Foundation Classes (IFC) models, which are externally provided data structures. This represents a potential surface for indirect prompt injection if malicious instructions are embedded in model metadata.\n
- Ingestion points: The
ifcLoader.load()method inSKILL.mdandreferences/examples.mdingests external model data.\n - Boundary markers: Not present in the provided code examples.\n
- Capability inventory: The skill uses capabilities for 3D rendering (
SimpleRenderer), model loading (IfcLoader), and metadata classification (Classifier).\n - Sanitization: No specific data sanitization or instruction-stripping for model metadata is described in these architectural guides.
Audit Metadata