The Model Context Protocol (MCP) that connects Large Language Models (LLMs) to external data and tools in ecosystems brings more context to AI chats and improves the answers that those models put out. But like nearly everything AI, MCPs have proliferated, outpacing security.