pdf2tex

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text from the user-supplied PDF is extracted at runtime via fitz.open(...); page.get_text("dict")["blocks"] and then used to generate LaTeX (e.g., section headings, table/cell text, bibliography), so the LLM context can include arbitrary text originating in an outsider document.