ncbi-pmc-skill

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

• Third-party content exposure detected (high risk: 0.70). The skill's required runtime script scripts/ncbi_pmc.py performs requests.get to the public NCBI PMC Open Access endpoint (https://www.ncbi.nlm.nih.gov/pmc/utils/oa/oa.fcgi) and parses the returned XML as part of its normal workflow (per SKILL.md), so untrusted public content from PMC is ingested and can influence the agent's summaries/responses.