google-shopping-search

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

• Third-party content exposure detected (high risk: 0.85). The script sends a user-provided q to Frevana and returns the API response JSON; this response is outsider-authored free text (public web/API content) that is read as text and then injected into the agent context via stdout (cat "$RESULT_FILE"), after only JSON validation.