latex-paper-en

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

• Third-party content exposure detected (high risk: 0.80). The skill's citation-verification workflow explicitly instructs programmatic searches and fetches from public sources (Semantic Scholar, CrossRef/doi.org, arXiv and Exa MCP) — see references/CITATION_VERIFICATION.md and the bibliography module references — so the agent will ingest untrusted third-party content that can affect citation/verification decisions.