We have talked a lot about how powerful and useful the Model Context Protocol is in the first parts of this book. We now know how to make servers with many features and smart clients, which lets us make apps that can think, act, and interact with the world in ways that have never been done before. But this power comes with a whole new set of very serious risks. In the "gold rush" phase of agentic development, big companies like Google, Microsoft, AWS, and Salesforce are all trying to get MCP-enabled products out as quickly as possible. In this rush, security is often overlooked, and the results are already making headlines.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

The Agentic Threat Landscape

  • Srinivasan Sekar

摘要

We have talked a lot about how powerful and useful the Model Context Protocol is in the first parts of this book. We now know how to make servers with many features and smart clients, which lets us make apps that can think, act, and interact with the world in ways that have never been done before. But this power comes with a whole new set of very serious risks. In the "gold rush" phase of agentic development, big companies like Google, Microsoft, AWS, and Salesforce are all trying to get MCP-enabled products out as quickly as possible. In this rush, security is often overlooked, and the results are already making headlines.