xlsx

Pass

Audited by Gen Agent Trust Hub on Aug 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The recalc.py script uses the subprocess module to execute soffice (LibreOffice) and system timeout utilities to recalculate Excel formulas. This is a functional requirement for the skill's operation.\n- [DYNAMIC_EXECUTION]: The recalc.py script generates a LibreOffice Basic macro (Module1.xba) and writes it to the user's application configuration directory at runtime to facilitate automation. This constitutes runtime generation and execution of code.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external spreadsheet files without explicit sanitization, which could lead to the agent following malicious instructions embedded within cell content.\n
  • Ingestion points: Data is loaded from files using pandas.read_excel() and openpyxl.load_workbook() in the SKILL.md examples and recalc.py script.\n
  • Boundary markers: None provided in the instructions or code.\n
  • Capability inventory: The skill has the ability to write files to the filesystem and execute shell commands through the LibreOffice integration.\n
  • Sanitization: No validation, filtering, or escaping is performed on spreadsheet data before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 21, 2026, 05:04 AM
Security Audit — agent-trust-hub — xlsx