docx

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill describes workflows for processing external .docx files and filling templates with data (‘SKILL.md’), which creates an inherent attack surface for indirect prompt injection if documents contain malicious instructions.
  • Ingestion points: Loading existing Word documents and filling templates using variables.
  • Boundary markers: No specific boundary delimiters or isolation instructions are provided for document content.
  • Capability inventory: The libraries mentioned (‘python-docx’, ‘docxtemplater’) facilitate reading and writing to the file system.
  • Sanitization: No explicit procedures for sanitizing or validating document content are mentioned in the best practices.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 08:05 AM
Security Audit — agent-trust-hub — docx