paper-2-web

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

• Third-party content exposure detected (high risk: 0.70). The skill can fetch and ingest open web content (e.g., "automatic logo discovery" via the Google Search API and citation/hover previews/links to DOIs and external sources that it downloads and integrates), so it clearly pulls untrusted third‑party web content into the agent's workflow.