polars-dovmed

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

• Third-party content exposure detected (high risk: 0.70). The skill explicitly fetches and ingests public PMC Open Access content (see SKILL.md instructions to use /api/search_literature, /api/scan_literature_advanced, and /api/get_paper_details against https://api.newlineages.com and the scripts/query_literature.py post_json call) and requires the agent to inspect abstracts/full_text and matched snippets as part of its workflow, exposing it to arbitrary third-party text that could influence subsequent decisions.