literature-search

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text can enter the LLM context via the fulltext/resolve path: tools/lit/sources/arxiv.py fetches arXiv HTML/LaTeX/PDF content from the public web at runtime (e.g., get_bytes(.../e-print/...) and latex_sections()), and the agent then reads that extracted text (or the generated arxiv_*.txt file) as prose for downstream LLM use.