ai-rag-pipeline

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

• Third-party content exposure detected (high risk: 0.90). The required workflow uses runtime web-search/extraction apps (e.g., tavily/search-assistant / tavily/extract) whose fetched outsider-authored page content or search results are then interpolated into the LLM prompt (e.g., $SEARCH_RESULT, $CONTENT) and sent to openrouter/*, creating an indirect prompt-injection path.