docx

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes local Python scripts such as unpack.py, pack.py, validate.py, and soffice.py to handle Word document structure and perform validation tasks.
  • [EXTERNAL_DOWNLOADS]: The skill references external tools and libraries including the docx-js npm package and pandoc for document creation and conversion processes.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes content from user-provided .docx files.
  • Ingestion points: Untrusted content is ingested from Word documents during reading and editing phases.
  • Boundary markers: No explicit delimiters are used to isolate document content from the agent's instructions.
  • Capability inventory: The skill possesses the capability to execute shell commands for file manipulation.
  • Sanitization: The skill implements structural validation via XSD schema checks (validate.py) to ensure the integrity of the generated OOXML output.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 01:25 AM
Security Audit — agent-trust-hub — docx