qinyan-citation

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

• Third-party content exposure detected (high risk: 0.70). The skill's scripts/search.sh and SKILL.md explicitly call the QINYAN OpenAPI to retrieve results from public third‑party sources (Google Scholar, Wanfang, PubMed, ArXiv) and require the agent to extract and act on that returned metadata to generate citations, so untrusted public content is ingested and can materially influence the agent's outputs.