clash-detection-analysis

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill performs geometric analysis on BIM models using established libraries such as ifcopenshell, numpy, and scipy. All operations are performed locally on user-provided data.
  • [PROMPT_INJECTION]: Analyzed the skill for indirect prompt injection surfaces. The skill ingests BIM element data (names and IDs) and renders them into HTML/Excel reports. 1. Ingestion points: Processes elements from IFC files via ifcopenshell.open() in SKILL.md. 2. Boundary markers: No specific boundary markers are used to separate element metadata from report templates. 3. Capability inventory: File system write access is used to export results in export_clash_report and generate_clash_html_report. 4. Sanitization: Uses standard pandas.to_html() for reporting. While this constitutes a data ingestion surface, there is no evidence of malicious intent or bypass attempts.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 04:28 AM