docx

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses the defusedxml library for all XML parsing tasks, which is a security best practice to prevent XML External Entity (XXE) attacks. This is implemented in scripts/document.py and scripts/utilities.py.
  • [COMMAND_EXECUTION]: The skill performs command execution for document conversion and validation using standard tools like pandoc, soffice (LibreOffice), and git. These executions are performed securely using subprocess.run with a list of arguments and shell=False, preventing command injection vulnerabilities.
  • [EXTERNAL_DOWNLOADS]: No suspicious runtime external downloads were detected. The skill references standard dependencies from trusted registries (APT, PyPI, NPM) in its documentation for environment setup.
  • [PROMPT_INJECTION]: The instructions within SKILL.md are focused on document structure and redlining workflows. No patterns of prompt injection, safety guideline bypass, or behavior overrides were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 01:45 PM