MCP
What is MCP
MCP (Model Context Protocol) is an open-source protocol developed by Anthropic, designed to standardize how AI applications provide context to large language models (LLMs). Think of MCP as the USB-C interface for AI applications—just as USB-C offers a standardized way for devices to connect various peripherals and accessories, MCP provides a standardized approach for AI models to interface with different data sources and tools.
The core of MCP adopts a Client-Server architecture, where an application can connect to multiple MCP Servers.
Once an AI application implements MCP, it can integrate with any MCP Server to extend its capabilities. After joining an MCP Server, the tool invocation process is as follows:
In this way, MCP achieves:
- Plug and Play: AI applications only need to implement the MCP protocol to access a rich ecosystem of third-party tools.
- Standardization: All tools adhere to a unified description format and invocation method.
- Decoupling Tools and Applications: Tool providers can independently develop and maintain tools without needing to understand the internal implementation of AI applications.
- Resource Sharing: Tools developed once can be reused by multiple AI applications.
How to use MCP
CloudBase implements the MCP mechanism for Agents. Agents can connect to the MCP Server deployed in a cloud-hosted environment to access its extended capabilities.
- Go to the Cloud Development Platform to create a new MCP Server
- Use MCP in Agent code
- Start a conversation. When the large language model (LLM) determines that a tool needs to be used, it will be automatically invoked.
For additional usage instructions, please see: