office-automation
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. The scripts perform their stated functions using standard file I/O operations and do not include any dangerous command execution or network exfiltration logic.\n- [PROMPT_INJECTION]: The skill processes content from external Word and Excel files, which represents an indirect prompt injection surface. However, no malicious instructions are present in the provided examples and the risk is associated with the primary skill purpose.\n
- Ingestion points:
scripts/word_processor.py(read_document),scripts/excel_processor.py(read_excel), andscripts/batch_processor.py(json.load).\n - Boundary markers: None (Uses standard string manipulation for template replacement).\n
- Capability inventory: File system read/write access for document manipulation and conversion.\n
- Sanitization: No specific sanitization of document text content before processing.\n- [EXTERNAL_DOWNLOADS]: Dependencies are limited to reputable, well-known Python packages (
python-docx,openpyxl,pandas) installed from the official Python Package Index (PyPI).
Audit Metadata