read-word

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill operates locally on the user's file system, reading document content only when explicitly requested. It does not perform any network requests or data exfiltration.
  • [SAFE]: The dependencies python-docx and olefile are well-known, established libraries for their respective formats and are used according to standard practices.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes content from external files, which presents a surface for indirect prompt injection.
  • Ingestion points: Text content is ingested from user-provided .docx and .doc files via the read_word_document function in read_word.py.
  • Boundary markers: No delimiters or safety instructions are used to separate the document content from the agent's internal instructions.
  • Capability inventory: The skill has file system access for reading documents and writing text output but lacks network or arbitrary command execution capabilities.
  • Sanitization: Content is extracted and passed to the agent as raw text strings without filtering or escaping techniques.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 01:02 AM
Security Audit — agent-trust-hub — read-word