docx

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes system binaries including pandoc for text extraction and LibreOffice (via a local Python wrapper) for document conversion. It also provides scripts for unpacking and repacking the XML structure of DOCX files.
  • [EXTERNAL_DOWNLOADS]: Includes instructions to install the 'docx' Node.js package from the official npm registry to support document generation.
  • [PROMPT_INJECTION]: Features a surface for indirect prompt injection as it ingests content from external Word documents. Evidence Chain: 1. Ingestion point: Reads .docx files using pandoc or unpack.py in SKILL.md. 2. Boundary markers: None present. 3. Capability inventory: Executes subprocesses for document tools in SKILL.md. 4. Sanitization: No explicit sanitization of document content before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 03:22 PM
Security Audit — agent-trust-hub — docx