docx
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill executes external command-line tools to perform document conversions and comparisons.
- ooxml/scripts/pack.py uses subprocess.run to call soffice (LibreOffice) for converting documents to HTML to validate their integrity.
- ooxml/scripts/validation/redlining.py uses subprocess.run to execute git diff for generating word-level differences between document versions.
- SKILL.md instructions direct the agent to use pandoc for text extraction and pdftoppm for image conversion.
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from Word documents, representing a potential surface for indirect prompt injection.
- Ingestion points: External document content is ingested through pandoc output or by reading XML files directly.
- Boundary markers: No specific delimiters are defined in the instructions for separating document content from instructions.
- Capability inventory: The skill has access to subprocess execution and file system operations.
- Sanitization: The skill mitigates XML-related risks by using the defusedxml library for parsing and manipulation.
- [EXTERNAL_DOWNLOADS]: The skill depends on external software and libraries.
- SKILL.md lists pandoc, LibreOffice, and poppler as required system-level dependencies.
- package.json specifies the docx Node.js package.
Audit Metadata