skills/openaec-foundation/blender-bonsai-ifcopenshell-sverchok-claude-skill-package/bonsai-impl-classification/Gen Agent Trust Hub
bonsai-impl-classification
Pass
Audited by Gen Agent Trust Hub on Mar 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: The skill focuses on domain-specific architectural data management tasks. No malicious patterns, obfuscation, or unauthorized data access were identified.
- [EXTERNAL_DOWNLOADS]: Fetches classification systems and property definitions from the buildingSMART Data Dictionary (bSDD) official API (api.bsdd.buildingsmart.org). This is a well-known industry-standard service.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface through the ingestion of external data from the bSDD API to classify BIM elements. However, the risk is negligible given the trusted nature of the buildingSMART infrastructure and the application of data to specific IFC attributes. Ingestion points: bsdd.Client.search_class and bsdd.Client.get_class in SKILL.md and references/examples.md. Boundary markers: None mentioned for external data. Capability inventory: Use of ifcopenshell.api.run for model modifications, model.write for file persistence, and bpy.ops.bim for Blender integration. Sanitization: Data is used to populate IFC entity properties without intermediate processing.
Audit Metadata