skills/datadrivenconstruction/ddc_skills_for_ai_agents_in_construction/bim-classification-ai/Gen Agent Trust Hub
bim-classification-ai
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implementation is transparent, using standard Python logic to map BIM element properties to classification standards like UniFormat and MasterFormat. No malicious patterns such as obfuscation, credential theft, or unauthorized network activity were detected.
- [DATA_EXPOSURE]: The skill requests 'filesystem' permission, which is appropriate for its function of processing project data files (Excel, CSV, JSON) and exporting classification results. The code does not attempt to access sensitive system files.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external data in the form of BIM element properties. This attack surface is mitigated by the implementation's use of regular expressions (re.findall) to tokenize input strings into words, effectively sanitizing the data before it is used for keyword matching and classification scoring.
Audit Metadata