ml-paper-writing

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

• Third-party content exposure detected (high risk: 0.90). The skill's required workflows explicitly instruct the agent to fetch and ingest open/public third‑party content (e.g., "Citation Verification (MANDATORY): Use WebSearch to verify EVERY citation", "Step 2: Search arXiv" and the paper‑miner flow "Analyze structure of https://arxiv.org/abs/…")—so untrusted web pages, arXiv/Google‑Scholar/Semantic Scholar/CrossRef results and arbitrary URLs are programmatically read and used to verify citations and update the knowledge base, which can materially influence the agent's decisions and tool use.