基于SpringAI 1.0.0 实现MCP

Lin_qx
342 阅读1分钟

MCP架构

MCP中文文档

image.png

  • MCP Hosts: 如 Claude Desktop、IDE 或 AI 工具，希望通过 MCP 访问数据的程序
  • MCP Clients: 维护与服务器一对一连接的协议客户端
  • MCP Servers: 轻量级程序，通过标准的 Model Context Protocol 提供特定能力
  • 本地数据源: MCP 服务器可安全访问的计算机文件、数据库和服务
  • 远程服务: MCP 服务器可连接的互联网上的外部系统（如通过 APIs）

数据流

  1. Prompt到达client
  2. client请求server，server选择合适的Function返回给client（可以根据需要，写自己的Function）
  3. client和model交互
  4. 返回结果

Server

  • 依赖
<parent>
    <groupId>org.springframework.boot</groupId>
    <artifactId>spring-boot-starter-parent</artifactId>
    <version>3.4.5</version>
    <relativePath/>
    <!-- lookup parent from repository -->
</parent>

<properties>
    <maven.compiler.source>17</maven.compiler.source>
    <maven.compiler.target>17</maven.compiler.target>
    <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
    <java.version>17</java.version>
    <spring-ai.version>1.0.0</spring-ai.version>

</properties>
<dependencies>
    <dependency>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-web</artifactId>
    </dependency>
    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-model-ollama</artifactId>
    </dependency>
    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-model-openai</artifactId>
    </dependency>

    <!--       MCP-->
    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-mcp-server</artifactId>
    </dependency>

    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-mcp-server-webmvc</artifactId>
    </dependency>

    <dependency>
        <groupId>org.projectlombok</groupId>
        <artifactId>lombok</artifactId>
        <optional>true</optional>
    </dependency>
    <dependency>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-test</artifactId>
        <scope>test</scope>
    </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>
  • 配置文件
# 端口
server.port=8080
# 不用但是要写，不然会报错
spring.ai.openai.base-url=https://api.gptsapi.net
spring.ai.openai.api-key=@{ollama.api-key}
spring.ai.openai.chat.options.model=gpt-3.5-turbo
# MCP 服务
spring.ai.mcp.server.name=music-mcp-server
#后续需要使用这个地址
spring.ai.mcp.server.sse-message-endpoint= /mcp/recommend
spring.ai.mcp.server.type = async
  • 对应的Tool
public interface RecommendService {
    String recommendInfo();
    String recommendJpop();
}

@Service
public class RecommendServiceImpl implements RecommendService {
    @Tool(description = "推荐中国风音乐")
    @Override
    public String recommendInfo() {
        return "推荐音乐人河图，河山万里，图一个你!";
    }
    @Tool(description = "推荐日本音乐")
    @Override
    public String recommendJpop() {
        return "JPOP领军人物米津玄师，你值得拥有!";
    }
}
  • 将你写的Tool注册给Spring
@Configuration
public class ToolCallbackProviderConfig {
    @Bean
    public ToolCallbackProvider recommendTools(RecommendService recommendService) {
        return MethodToolCallbackProvider.builder().toolObjects(recommendService).build();
    }
}

Client

  • 依赖
<parent>
    <groupId>org.springframework.boot</groupId>
    <artifactId>spring-boot-starter-parent</artifactId>
    <version>3.4.5</version>
    <relativePath/>
    <!-- lookup parent from repository -->
</parent>

<properties>
    <maven.compiler.source>17</maven.compiler.source>
    <maven.compiler.target>17</maven.compiler.target>
    <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
    <java.version>17</java.version>
    <spring-ai.version>1.0.0</spring-ai.version>

</properties>
<dependencies>
    <dependency>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-web</artifactId>
    </dependency>
    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-model-ollama</artifactId>
    </dependency>
    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-model-openai</artifactId>
    </dependency>

    <!--       MCP-->
    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-autoconfigure-mcp-client</artifactId>
    </dependency>
    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-mcp-client-webflux</artifactId>
    </dependency>

    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-mcp-server</artifactId>
    </dependency>

    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-mcp-server-webmvc</artifactId>
    </dependency>

    <dependency>
        <groupId>org.projectlombok</groupId>
        <artifactId>lombok</artifactId>
        <optional>true</optional>
    </dependency>
    <dependency>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-test</artifactId>
        <scope>test</scope>
    </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>
  • 配置文件
spring.application.name=mcp-client
# 随便写什么都行，不然会启动失败
spring.ai.openai.api-key=your_openai_api_key_here
#  ollama
spring.ai.ollama.base-url=http://localhost:11434
#需要选择支持tool的模型
spring.ai.ollama.chat.model=qwen3

spring.mvc.async.request-timeout=0
spring.codec.charset=UTF-8
spring.codec.max-in-memory-size=1MB


# 应用服务 WEB 访问端口
server.port=8081

# MCP客户端的配置
spring.ai.mcp.client.name=mcp-client
# 要连接的服务地址
spring.ai.mcp.client.sse.connections.server1.url=http://localhost:8080
# 默认情况下，工具回调功能处于禁用状态，需要开启,否则启动报错
spring.ai.mcp.client.toolcallback.enabled = true
  • Controller
@RestController
public class McpController {
    @Autowired
    private OllamaChatModel ollamaChatModel;
    @Autowired
    private ToolCallbackProvider toolCallbackProvider;

    @GetMapping("/mcp/generate")
    public String generate(@RequestParam(value = "message", defaultValue = "推荐一首歌") String message) {
        ChatClient chatClient = ChatClient.builder(ollamaChatModel)
                .defaultToolCallbacks(toolCallbackProvider.getToolCallbacks())
                .build();
        ChatClient.CallResponseSpec call = chatClient.prompt(message).call();
        return call.content();

    }
}

测试

avatar
文章
阅读
粉丝