SpringAI2.0MCP的简单使用

模型上下文协议（MCP）是一种标准化协议，使 AI 模型能够以结构化的方式与外部工具和资源进行交互。您可以将其视为连接您的 AI 模型与现实世界的桥梁——通过一致的接口，让它们访问数据库、API、文件系统及其他外部服务。它支持多种传输机制，可在不同环境中提供灵活性。

SpringAI2.0提供了MCP的服务端实现和客户端实现，这里要用两个项目

一、服务端

1、依赖pom.xml

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 https://maven.apache.org/xsd/maven-4.0.0.xsd">
    <modelVersion>4.0.0</modelVersion>
    <parent>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-parent</artifactId>
        <version>4.1.2</version>
        <relativePath/> <!-- lookup parent from repository -->
    </parent>
    <groupId>com.suibibk</groupId>
    <artifactId>spring-ai-test</artifactId>
    <version>0.0.1-SNAPSHOT</version>
    <name>spring-ai-test</name>
    <description>spring-ai-test</description>
    <properties>
        <java.version>17</java.version>
        <spring-ai.version>2.0.1</spring-ai.version>
    </properties>
    <dependencies>
        <!-- MCP Server  支持：`STREAMABLE` / `SSE` / `STATELESS`（http 网络模式，复用 8080 端口，访问 `/mcp`）-->
        <dependency>
            <groupId>org.springframework.ai</groupId>
            <artifactId>spring-ai-starter-mcp-server-webmvc</artifactId>
        </dependency>
    </dependencies>
    <dependencyManagement>
        <dependencies>
            <dependency>
                <groupId>org.springframework.ai</groupId>
                <artifactId>spring-ai-bom</artifactId>
                <version>${spring-ai.version}</version>
                <type>pom</type>
                <scope>import</scope>
            </dependency>
        </dependencies>
    </dependencyManagement>
    <build>
        <plugins>
            <plugin>
                <groupId>org.springframework.boot</groupId>
                <artifactId>spring-boot-maven-plugin</artifactId>
            </plugin>
        </plugins>
    </build>
</project>

2、配置application.yaml

spring:
  application:
    name: spring-ai-mcp-test
  ai:
    mcp:
      server:
        protocol: STREAMABLE
logging:
  level:
    org.springframework.ai: debug
server:
  port: 8081

3、mcp工具类

@Component
public class CalculatorTools {
    private final Map<String, String> configData = Map.of(
            "name", "demo",
            "version", "1.0.0"
    );
    @McpTool(name = "add", description = "Add two numbers together")
    public int add(
            @McpToolParam(description = "First number", required = true) int a,
            @McpToolParam(description = "Second number", required = true) int b) {
        System.out.println("有调用："+a+";"+b);
        return a + b;
    }
    @McpResource(uri = "config://{key}", name = "Configuration", description = "Read config value by key")
    public String getConfig(String key) {
        return configData.getOrDefault(key, "not found");
    }
}

4、启动类

@SpringBootApplication
public class SpringAiMCPTestApplication {
    public static void main(String[] args) {
        SpringApplication.run(SpringAiMCPTestApplication.class, args);
    }
}

然后启动就好了。

二、客户端

客户端就基于之前的例子，很简单

1、依赖

 <!-- MCP Client 依赖，服务端包不要引进来 -->
        <dependency>
            <groupId>org.springframework.ai</groupId>
            <artifactId>spring-ai-starter-mcp-client</artifactId>
        </dependency>

2、配置

spring:
  application:
    name: spring-ai-test
  ai:
    openai:
      api-key: ${AI_BAILIAN_API_KEY}
      base-url: https://dashscope.aliyuncs.com/compatible-mode/v1
      chat:
        model: qwen3.8-flash
        temperature: 0.7
      embedding:
        model: qwen3.7-text-embedding-flash
        dimensions: 2048
    mcp:
      client:
        streamable-http:
          connections:
            self-server:
              url: http://localhost:8081
logging:
  level:
    org.springframework.ai: debug

3、调用

 @Autowired
    private  ToolCallbackProvider mcpTools;
   @GetMapping("/mcp")
    public String mcp(String topic){
        String result =  openaiChatClient
                .prompt()
                .user(u -> u.text("{topic}").param("topic",topic))
                .tools(mcpTools)
                .call()
                .content();
        return result;
    }

4、测试

访问http://localhost:8080/mcp?topic=1加上1等于多少
结果返回1 加上 1 等于 2。看MCPserver的日志也打印了

有调用：1;1