editing-word-documents

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow is to read/analyze a user-supplied .docx (e.g., via pandoc document.docx -o output.md or python-docx/raw XML extraction), and the document body/comments XML can contain outsider-authored free text that becomes LLM-readable context after conversion/parsing.