fact-checker

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow explicitly requires accessing and evaluating external sources—steps "Verify existence" and "Verify content" plus the Integration section's instruction to use pubmed-database/openalex-database and to check DOIs—meaning the agent will fetch public third‑party URLs/papers and use their content to drive verification decisions.