research-assistant

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

• Third-party content exposure detected (high risk: 0.90). The skill's core reference prompts explicitly require searching for and ingesting open/public sources (e.g., CNKI/Google Scholar/PubMed/website searches and using tools like browse_page or uploading articles to Gemini) as part of normal workflows — see references/01-literature-retrieval.md (文献检索 prompts), references/32 (Canvas model: "从CNKI、Google Scholar、JSTOR、PubMed等数据库搜索"), and references/04-tech-roadmap.md (instructions to upload articles to Gemini) — so it clearly exposes the agent to untrusted third‑party content that could carry indirect prompt injection.