word-documents

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The scripts analyze.py, comments.py, compare.py, create.py, format.py, merge.py, and track_changes.py automatically install python-docx, lxml, and docxcompose from PyPI using pip if they are missing at runtime.\n- [COMMAND_EXECUTION]: The convert.py script executes the pandoc system binary via the subprocess module to handle document conversions, allowing for the inclusion of arbitrary extra arguments provided by the user.\n- [PROMPT_INJECTION]: The skill's ability to process untrusted external files (.docx, .md, .html, .json) creates a surface for indirect prompt injection attacks.\n
  • Ingestion points: Input files provided to processing scripts, specifically those parsed by the Document objects in the scripts folder.\n
  • Boundary markers: There are no boundary markers or instructions used to isolate external content from the agent's internal logic.\n
  • Capability inventory: The skill possesses the capability to perform file system operations (read/write), execute external processes (pandoc), and install software packages.\n
  • Sanitization: The skill does not perform any sanitization, filtering, or validation of content extracted from processed documents before it is used in subsequent operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 07:31 AM