alterlab-literature-review

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md workflow explicitly requires searching and ingesting open/public third-party sources (e.g., "gget search pubmed", "gget search biorxiv", "Search arXiv via direct API", "Search Semantic Scholar via API", and even Google Scholar scraping") and the scripts (search_databases.py, verify_citations.py) fetch and process remote metadata, so untrusted web content is read and used to drive screening, synthesis, and follow-up actions—allowing malicious or crafted third-party content to influence agent decisions.