skills/yamato-snow/skills/xlsx/Gen Agent Trust Hub

xlsx

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION] (LOW): The script executes the soffice (LibreOffice) binary and system utilities like timeout or gtimeout using subprocess.run. The implementation uses a list of arguments rather than a raw string with shell=True, which effectively prevents command injection vulnerabilities from malicious filenames.
  • [DYNAMIC_EXECUTION] (LOW): The script generates a StarBasic macro file (Module1.xba) within the user's LibreOffice configuration directory (~/.config/libreoffice or ~/Library/Application Support/LibreOffice). This allows the script to trigger a 'calculate all' and 'save' operation within the headless LibreOffice process. While this modifies the local environment, the code being written is hardcoded and safe.
  • [INDIRECT_PROMPT_INJECTION] (LOW): The skill processes untrusted Excel files provided as input.
  • Ingestion points: recalc.py reads Excel files using openpyxl.load_workbook and LibreOffice.
  • Boundary markers: None. The script iterates through all cells to find error strings.
  • Capability inventory: File system write (macro installation) and subprocess execution (soffice).
  • Sanitization: No specific sanitization is performed on the Excel content, though the script only looks for specific error tokens like #VALUE! or #DIV/0!, which limits the exploit surface.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:49 PM