docx

Pass

Audited by Gen Agent Trust Hub on Jun 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses subprocess.run to execute soffice (LibreOffice) for document validation and git diff for comparing redlined versions of documents. These operations are constrained to specific tasks and operate on files within temporary or agent-controlled directories.
  • [EXTERNAL_DOWNLOADS]: The documentation lists several well-known and trusted external dependencies required for document processing, including pandoc, LibreOffice, the docx Node.js library, and defusedxml. These are standard tools in the document processing ecosystem.
  • [PROMPT_INJECTION]: The skill processes user-supplied .docx files, which inherently presents a surface for indirect prompt injection if those documents contain adversarial instructions. This risk is common to all document-analysis skills and is managed here through the use of secure parsing and standard agent logic.
  • [SAFE]: The skill explicitly utilizes defusedxml to handle XML content, demonstrating a security-first approach to processing Office Open XML files. No obfuscation, data exfiltration, or persistence mechanisms were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 20, 2026, 10:22 PM
Security Audit — agent-trust-hub — docx