Your LangChain Agent Has No Security. Neither Does CrewAI, OpenAI, or 6 Others.

Every AI framework lets your agent call tools. None of them check what those tools actually...