paper-writing

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly fetches and interprets external web content — e.g., the optional --style-ref accepts http(s) URLs processed by tools/extract_paper_style.py and the citation-audit (Phase 5.8 / Phase 6 pre‑flight) performs web/DBLP/arXiv lookups whose findings are read and used to modify citations, block the final report, and drive follow-up actions — thus exposing the agent to untrusted third‑party content that can materially influence behavior.