synthesis-fact-checking

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

• Third-party content exposure detected (high risk: 0.80). This skill's SKILL.md explicitly instructs the agent to locate and verify claims by searching and reading public sources (e.g., "search for the exact quote", "Use Google Scholar, the publisher's website, or the institution's publications page", and "Search Retraction Watch" and social-media posts), so the agent is expected to ingest and act on untrusted third‑party web/user-generated content as part of its workflow.