google-news-search

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

• Third-party content exposure detected (high risk: 0.85). The skill calls an external Google News search API via curl and then reads the returned HTTP response body from RESPONSE_FILE into python3 for JSON validation and finally prints it to stdout; that response is outsider-authored free text (news content) that becomes LLM-readable context.