ontology-stack-builder

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

• Third-party content exposure detected (medium risk: 0.30). Outsider-authored free text can enter the required runtime via the user-provided “business questions” (ask_user_question in Phase 1, later embedded into the Phase 7 agent call JSON), and the workflow then runs SNOWFLAKE.CORTEX.AGENT(... PARSE_JSON({"messages":[{"role":"user","content":"<domain question from Phase 1>"}]})), making that text readable to the LLM at runtime.