google-events-search

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

• Third-party content exposure detected (high risk: 0.85). The skill sends the user-provided --q (free-form query text) to Frevana’s /service/google-events and then prints the returned SerpAPI JSON verbatim; that JSON is outsider-authored content fetched at runtime from a third-party API, which becomes LLM-readable context via the agent’s stdout ingestion.