multi-tenant-db

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow ingests and executes outsider-authored free text indirectly when a user asks database-routing/schema questions that the agent answers by calling MCP SQL tools, which accept LLM-provided search/query parameters against platform/tenant tables (e.g., Insightt/Foundd routing tables like platform_companies/client).