5、Spring AI（MCPServer+MCPClient+Ollama）開發環境搭建

前言：

該開發環境是在?3、后端持久化（SpringBoot3.5.0+MybatisPlus3.5.5+mysql8.4.0）環境搭建

上進行改造的，用到了后端持久化，主要改造的地方為數據庫把email字段改為height（身高），該開發環境主要是設計了一個灌籃高手籃球經理對球隊成員簡單的查詢，通過這個場景把MCPServer、MCPClient、大模型、用戶客戶端相互的職責和關系簡單捋一下，其他的改造和實現一邊貼代碼一邊解釋

1、數據庫表結構

數據庫表結構以及數據初始化、user_info表

ET NAMES utf8mb4;
SET FOREIGN_KEY_CHECKS = 0;-- ----------------------------
-- Table structure for user_info
-- ----------------------------
DROP TABLE IF EXISTS `user_info`;
CREATE TABLE `user_info`  (`id` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci NOT NULL COMMENT 'ID主鍵',`user_name` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci NULL DEFAULT NULL COMMENT '用戶名',`sex` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci NULL DEFAULT NULL COMMENT '性別',`hobby` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci NULL DEFAULT NULL COMMENT '愛好',`special_skill` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci NULL DEFAULT NULL COMMENT '特長',`score` int(0) NULL DEFAULT NULL COMMENT '評分',`height` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci NULL DEFAULT NULL COMMENT '身高',PRIMARY KEY (`id`) USING BTREE
) ENGINE = InnoDB CHARACTER SET = utf8mb4 COLLATE = utf8mb4_unicode_ci ROW_FORMAT = Dynamic;-- ----------------------------
-- Records of user_info
-- ----------------------------
INSERT INTO `user_info` VALUES ('3830a6dcd94e60e41e491a8701e10a48', '流川楓', '男', '睡覺', '大前鋒', 98, '187cm');
INSERT INTO `user_info` VALUES ('463097bd7d06c7fbf709c40207c356c3', '牧紳一', '男', '未知', '后衛', 98, '184cm');
INSERT INTO `user_info` VALUES ('552cef7ea26dcb134623891417527cff', '櫻木花道', '男', '赤木晴子', '大前鋒', 97, '188cm');
INSERT INTO `user_info` VALUES ('6ebee37985f6cdf0ff6e7520c5441616', '仙道彰', '男', '釣魚', '小前鋒', 98, '190cm');
INSERT INTO `user_info` VALUES ('716bd21e54e5964058e4b9ad78cbaca3', '宮城良田', '男', '井上彩子', '后衛', 96, '168cm');
INSERT INTO `user_info` VALUES ('821a92f51a87be2d5d052e10a7223e70', '三井壽', '男', '三分球', '小前鋒', 98, '184cm');
INSERT INTO `user_info` VALUES ('9f5553b0b67821bae5b5869fc8ff3a69', '赤木晴子', '女', '籃球迷', '啦啦隊', 60, '156cm');
INSERT INTO `user_info` VALUES ('af2ab75e939b5cff8a95569b306be911', '井上彩子', '女', '宮城良田', '籃球經理', 70, '163cm');
INSERT INTO `user_info` VALUES ('d3f967929558b8c834396dd8aaa24e24', '澤北榮治', '男', '打籃球', '小前鋒', 98, '186cm');
INSERT INTO `user_info` VALUES ('e6bf09db5717d9c60af41c9dde65f0d6', '赤木剛憲', '男', '吃香蕉', '中鋒', 97, '199cm');SET FOREIGN_KEY_CHECKS = 1;

該表中存儲了一些球員的基本信息，以便籃球經理查看

2、MCPServer搭建

（1）pom.xml文件

該文件在之前??3、后端持久化（SpringBoot3.5.0+MybatisPlus3.5.5+mysql8.4.0）環境搭建

的基礎上改動不多，整體如下：

<?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 http://maven.apache.org/xsd/maven-4.0.0.xsd"><modelVersion>4.0.0</modelVersion><groupId>com.everything.autotest</groupId><artifactId>spring_ai_project</artifactId><version>1.0-SNAPSHOT</version><properties><maven.compiler.source>17</maven.compiler.source><maven.compiler.target>17</maven.compiler.target><java.version>17</java.version><spring-boot.version>3.5.0</spring-boot.version><project.build.sourceEncoding>UTF-8</project.build.sourceEncoding></properties><parent><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-parent</artifactId><version>3.5.0</version></parent><dependencies><!-- mcp-server-webmvc --><dependency><groupId>org.springframework.ai</groupId><artifactId>spring-ai-starter-mcp-server-webmvc</artifactId><version>1.0.0</version></dependency><!-- Web 模塊 --><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-web</artifactId><exclusions><exclusion><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-logging</artifactId></exclusion></exclusions></dependency><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter</artifactId><exclusions><exclusion><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-logging</artifactId></exclusion></exclusions></dependency><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-test</artifactId><scope>test</scope></dependency><dependency><groupId>org.projectlombok</groupId><artifactId>lombok</artifactId><optional>true</optional></dependency><!-- MyBatisPlus --><dependency><groupId>com.baomidou</groupId><artifactId>mybatis-plus-spring-boot3-starter</artifactId><version>3.5.5</version></dependency><dependency><groupId>com.mysql</groupId><artifactId>mysql-connector-j</artifactId></dependency><!-- Selenium WebDriver --><dependency><groupId>org.seleniumhq.selenium</groupId><artifactId>selenium-java</artifactId><version>4.31.0</version></dependency><!-- WebDriver Manager --><dependency><groupId>io.github.bonigarcia</groupId><artifactId>webdrivermanager</artifactId><version>6.1.0</version></dependency><!-- SLF4J API --><dependency><groupId>org.slf4j</groupId><artifactId>slf4j-api</artifactId><version>2.0.9</version></dependency><!-- Log4j2核心實現包 --><dependency><groupId>org.apache.logging.log4j</groupId><artifactId>log4j-core</artifactId><version>2.24.3</version></dependency><!-- SLF4J與Log4j2的橋接器 --><dependency><groupId>org.apache.logging.log4j</groupId><artifactId>log4j-slf4j2-impl</artifactId><version>2.24.3</version><exclusions><exclusion><artifactId>log4j-core</artifactId><groupId>org.apache.logging.log4j</groupId></exclusion></exclusions></dependency></dependencies></project>

只增加了一個MCPServer的必須文件，注意版本號，使用當前最新版本Spring AI 1.0.0，如下：

<!-- mcp-server-webmvc -->
<dependency><groupId>org.springframework.ai</groupId><artifactId>spring-ai-starter-mcp-server-webmvc</artifactId><version>1.0.0</version>
</dependency>

（2）application.yml文件

文件加入了Spring AI所需要的最簡參數，整體application.yml如下：

server:port: 9091
spring:ai:mcp:server:name: slamdunk-mcp-serverversion: 1.0.0sse-endpoint: /sse/slamdunk-mcp-serverdatasource:url: jdbc:mysql://localhost:3306/everythingtest?useUnicode=true&characterEncoding=utf8&characterSetResults=utf8username: rootpassword: 1234driver-class-name: com.mysql.cj.jdbc.Driverdebug: true
mybatis-plus:mapper-locations: classpath:mapper/*.xmlglobal-config:db-config:id-type: ASSIGN_UUIDfield-strategy: NOT_EMPTYdb-type: MYSQLconfiguration:map-underscore-to-camel-case: truecall-setters-on-nulls: truelog-impl: org.apache.ibatis.logging.stdout.StdOutImpl

MCPServer的yml文件配置唯一注意的一點就是sse-endpoint: /sse/slamdunk-mcp-server

sse-endpoint這個參數需要在代碼里做一些特殊處理才能生效，處理完成后MCPClient調用時才不會報404錯誤

（3）java文件

UserInfoMCPServerFacade.java

package com.everything.autotest.demo.mcpserverdemo;import com.baomidou.mybatisplus.core.conditions.query.QueryWrapper;
import com.everything.autotest.demo.mybatisplus3demo.UserInfo;
import com.everything.autotest.demo.mybatisplus3demo.userinfo.UserInfoService;
import org.springframework.ai.tool.annotation.Tool;
import org.springframework.ai.tool.annotation.ToolParam;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Service;import java.util.List;@Service
public class UserInfoMCPServerFacade {@AutowiredUserInfoService userInfoService;@Tool(name = "getInfo", description = "根據球員姓名獲取某個球員的詳細信息")public UserInfo getInfo(@ToolParam(description = "球員姓名") String userName) throws Exception {var userInfoQueryWrapper = new QueryWrapper<UserInfo>();userInfoQueryWrapper.like("user_name", userName);return userInfoService.getOne(userInfoQueryWrapper);}@Tool(name = "getList", description = "獲取所有球員信息")public List<UserInfo> getList() {return userInfoService.list();}
}

做一個簡單解釋：

1、簡單的可以理解為這個類相當于之前的Controller層，之前的Controller層的參數是從頁面或者第三方的restful調用過來，現在改為從MCPClient客戶端調用；調過來之后，后面處理的邏輯不變，依然是調用UserInfoService層來查詢數據庫和處理邏輯并返回到UserInfoMCPServerFacade

2、為什么這個類名后綴起成Facade，因為目前也沒啥標準規范，個人理解是，如果起名成Controller感覺有點怪怪的，如果起名成Service又和對應持久化層的Service有點混淆，所以干脆起成門面（Facade），這樣也有面子。

3、關于注解，及api使用方法會在后面更細節的分析，這篇先把運行環境搭起來

SlamDunkMCPRegisterConfiguration.java

package com.everything.autotest.demo.mcpserverdemo;import org.springframework.ai.tool.ToolCallbackProvider;
import org.springframework.ai.tool.method.MethodToolCallbackProvider;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;@Configuration
public class SlamDunkMCPRegisterConfiguration {@Beanpublic ToolCallbackProvider userInfoTools(UserInfoMCPServerFacade userInfoMCPServerFacade){return MethodToolCallbackProvider.builder().toolObjects(userInfoMCPServerFacade).build();}
}

把UserInfoMCPServerFacade搞進來

SlamDunkMCPServerConfig.java

package com.everything.autotest.demo.mcpserverdemo;import com.fasterxml.jackson.databind.ObjectMapper;
import io.modelcontextprotocol.server.transport.WebMvcSseServerTransportProvider;
import org.springframework.ai.mcp.server.autoconfigure.McpServerProperties;
import org.springframework.beans.factory.ObjectProvider;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
import org.springframework.context.annotation.Primary;
import org.springframework.web.servlet.function.RouterFunction;
import org.springframework.web.servlet.function.ServerResponse;@Configuration
public class SlamDunkMCPServerConfig {@Bean@Primarypublic WebMvcSseServerTransportProvider webMvcSseServerTransportProvider(ObjectProvider<ObjectMapper> objectMapperProvider, McpServerProperties serverProperties) {ObjectMapper objectMapper = objectMapperProvider.getIfAvailable(ObjectMapper::new);return new WebMvcSseServerTransportProvider(objectMapper, serverProperties.getSseMessageEndpoint(),serverProperties.getSseEndpoint());}@Beanpublic RouterFunction<ServerResponse> mvcMcpRouterFunction(WebMvcSseServerTransportProvider transportProvider) {return transportProvider.getRouterFunction();}}

SlamDunkMCPServerConfig.java這個類完全就是為了讓sse-endpoint這個參數生效，其實可以不配sse-endpoint這個參數也行，但是感覺配出來高大上一點；

通過以上配置MCPServer就搭建完畢了，運行起來，瀏覽器輸入

http://localhost:9091/sse/slamdunk-mcp-server

如果界面是這樣，說明mcpServer配置成功

MCPServer搭建成功

3、MCPClient搭建

MCPClient和MCPServer是兩個不同的工程，最好不要寫在一個工程里分兩個Model，這樣容易串端口

在這之前，首先參考??4、大模型本地運行環境搭建?把本地大模型搭起來

然后瀏覽器輸入?http://localhost:11434/?如果顯示下圖，則說明大模型運行成功

（1）pom.xml文件

文件必要依賴如下：

<!-- mcp-client -->
<dependency><groupId>org.springframework.ai</groupId><artifactId>spring-ai-starter-mcp-client</artifactId>
</dependency>
<!-- 使用的大模型依賴 -->
<dependency><groupId>org.springframework.ai</groupId><artifactId>spring-ai-starter-model-ollama</artifactId>
</dependency>

<dependencyManagement><dependencies><dependency><groupId>org.springframework.ai</groupId><artifactId>spring-ai-bom</artifactId><version>1.0.0</version><type>pom</type><scope>import</scope></dependency></dependencies>
</dependencyManagement>

以上配置我不確定能不能生效，因為對于Client配置我改了很多次，又沒有重新install可能會有緩存，如果不生效或者有問題，請參考Server的配置，把版本號顯示寫

還是要注意：Spring AI的版本號一定要選1.0.0版本，當前最新版

（2）application.yml文件

server:port: 9100
spring:ai:ollama:base-url: http://localhost:11434chat:model: llama3.2mcp:client:enabled: truename: slamdunk-mcp-clientversion: 1.0.0
#        request-timeout: 30stype: syncsse:connections:server1:url: http://localhost:9091sse-endpoint: /sse/slamdunk-mcp-servertoolcallback:enabled: true

這里的sse-endpoint: /sse/slamdunk-mcp-server與服務器保持一致（注意：一定要用1.0.0版，別用M版本）

（3）java文件

SlamDunkClientConfig.java

package com.gempharmatech.mcpclient;import org.springframework.ai.chat.client.ChatClient;
import org.springframework.ai.chat.client.advisor.SimpleLoggerAdvisor;
import org.springframework.ai.ollama.OllamaChatModel;
import org.springframework.ai.tool.ToolCallbackProvider;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;@Configuration
public class SlamDunkClientConfig {@Autowiredprivate OllamaChatModel ollamaChatModel;@Autowiredprivate ToolCallbackProvider toolCallbackProvider;@Beanpublic ChatClient chatClient() {StringBuilder sb = new StringBuilder();sb.append("你是一個籃球經理，需要對籃球隊內的一切事務所處判斷和決策。");sb.append("根據球員姓名獲取某個球員的詳細信息。");sb.append("獲取所有球員信息。");sb.append("最后，可以通過詼諧幽默的方式回答出問題，重點是詼諧幽默，可以允許你自由發揮");return ChatClient.builder(ollamaChatModel).defaultSystem(sb.toString()).defaultAdvisors(new SimpleLoggerAdvisor()).defaultToolCallbacks(toolCallbackProvider).build();}
}

SlamDunkMCPClintController.java?

package com.gempharmatech.mcpclient;import com.fasterxml.jackson.databind.ObjectMapper;
import jakarta.annotation.Resource;
import org.springframework.ai.chat.client.ChatClient;
import org.springframework.web.bind.annotation.PostMapping;
import org.springframework.web.bind.annotation.RequestBody;
import org.springframework.web.bind.annotation.RestController;@RestController
public class SlamDunkMCPClintController {@Resourceprivate ObjectMapper objectMapper;@Resourceprivate ChatClient chatClient;@PostMapping("/slamdunk/gogogo")public String getInfo(@RequestBody String jsonData) throws Exception {var node = objectMapper.readTree(jsonData);var message = node.get("message").asText();return chatClient.prompt().user(message).call().content();}
}