add

eino-project/api/customer/v1/customer.pb.go

@@ -1,6 +1,6 @@
 // Code generated by protoc-gen-go. DO NOT EDIT.
 // versions:
-// 	protoc-gen-go v1.36.9
+// 	protoc-gen-go v1.36.10
 // 	protoc        v3.21.12
 // source: customer/v1/customer.proto
 

eino-project/api/helloworld/v1/error_reason.pb.go

@@ -1,6 +1,6 @@
 // Code generated by protoc-gen-go. DO NOT EDIT.
 // versions:
-// 	protoc-gen-go v1.36.9
+// 	protoc-gen-go v1.36.10
 // 	protoc        v3.21.12
 // source: helloworld/v1/error_reason.proto
 

eino-project/api/helloworld/v1/greeter.pb.go

@@ -1,6 +1,6 @@
 // Code generated by protoc-gen-go. DO NOT EDIT.
 // versions:
-// 	protoc-gen-go v1.36.9
+// 	protoc-gen-go v1.36.10
 // 	protoc        v3.21.12
 // source: helloworld/v1/greeter.proto
 

eino-project/cmd/server/main.go

@@ -1,20 +1,19 @@
 package main
 
 import (
-	"flag"
+    "flag"
-	"os"
+    "os"
 
-	"eino-project/internal/conf"
+    "eino-project/internal/conf"
 
-	"github.com/go-kratos/kratos/v2"
+    "github.com/go-kratos/kratos/v2"
-	"github.com/go-kratos/kratos/v2/config"
+    "github.com/go-kratos/kratos/v2/config"
-	"github.com/go-kratos/kratos/v2/config/file"
+    "github.com/go-kratos/kratos/v2/config/file"
-	"github.com/go-kratos/kratos/v2/log"
+    "github.com/go-kratos/kratos/v2/log"
-	"github.com/go-kratos/kratos/v2/middleware/tracing"
+    "github.com/go-kratos/kratos/v2/middleware/tracing"
-	"github.com/go-kratos/kratos/v2/transport/grpc"
+    "github.com/go-kratos/kratos/v2/transport/http"
-	"github.com/go-kratos/kratos/v2/transport/http"
 
-	_ "go.uber.org/automaxprocs"
+    _ "go.uber.org/automaxprocs"
 )
 
 // go build -ldflags "-X main.Version=x.y.z"
@@ -33,18 +32,17 @@ func init() {
 	flag.StringVar(&flagconf, "conf", "./configs", "config path, eg: -conf config.yaml")
 }
 
-func newApp(logger log.Logger, gs *grpc.Server, hs *http.Server) *kratos.App {
+func newApp(logger log.Logger, hs *http.Server) *kratos.App {
-	return kratos.New(
+    return kratos.New(
-		kratos.ID(id),
+        kratos.ID(id),
-		kratos.Name(Name),
+        kratos.Name(Name),
-		kratos.Version(Version),
+        kratos.Version(Version),
-		kratos.Metadata(map[string]string{}),
+        kratos.Metadata(map[string]string{}),
-		kratos.Logger(logger),
+        kratos.Logger(logger),
-		kratos.Server(
+        kratos.Server(
-			gs,
+            hs,
-			hs,
+        ),
-		),
+    )
-	)
 }
 
 func main() {

eino-project/cmd/server/wire.go

@@ -6,17 +6,13 @@
 package main
 
 import (
-	"eino-project/internal/ai"
 	"eino-project/internal/biz"
 	"eino-project/internal/conf"
-	contextpkg "eino-project/internal/context"
 	"eino-project/internal/data"
 	"eino-project/internal/data/repoimpl"
-	"eino-project/internal/monitor"
+	"eino-project/internal/domain"
 	"eino-project/internal/server"
 	"eino-project/internal/service"
-	"eino-project/internal/session"
-	"eino-project/internal/vector"
 
 	"github.com/go-kratos/kratos/v2"
 	"github.com/go-kratos/kratos/v2/log"
@@ -25,5 +21,13 @@ import (
 
 // wireApp init kratos application.
 func wireApp(*conf.Server, *conf.Data, *conf.Bootstrap, log.Logger) (*kratos.App, func(), error) {
-    panic(wire.Build(server.ProviderSet, data.ProviderSet, repoimpl.ProviderSet, biz.ProviderSet, service.ProviderSet, ai.ProviderSet, session.ProviderSet, vector.ProviderSet, monitor.ProviderSet, contextpkg.ProviderSet, newApp))
+	panic(wire.Build(
+		server.ProviderSet,
+		data.ProviderSet,
+		repoimpl.ProviderSet,
+		biz.ProviderSet,
+		service.ProviderSet,
+		domain.ProviderSet,
+		newApp,
+	))
 }

eino-project/cmd/server/wire_gen.go

@@ -7,17 +7,16 @@
 package main
 
 import (
-	"eino-project/internal/ai"
 	"eino-project/internal/biz"
 	"eino-project/internal/conf"
-	"eino-project/internal/context"
 	"eino-project/internal/data"
 	"eino-project/internal/data/repoimpl"
-	"eino-project/internal/monitor"
+	"eino-project/internal/domain/context"
+	"eino-project/internal/domain/monitor"
+	"eino-project/internal/domain/session"
+	"eino-project/internal/domain/vector"
 	"eino-project/internal/server"
 	"eino-project/internal/service"
-	"eino-project/internal/session"
-	"eino-project/internal/vector"
 	"github.com/go-kratos/kratos/v2"
 	"github.com/go-kratos/kratos/v2/log"
 )
@@ -36,7 +35,7 @@ func wireApp(confServer *conf.Server, confData *conf.Data, bootstrap *conf.Boots
 	}
 	documentProcessor := vector.NewDocumentProcessor(vectorService)
 	knowledgeSearcher := vector.NewKnowledgeSearcher(vectorService)
-	contextManager := context.NewContextManagerFromBootstrapConfig(bootstrap, logger)
+	contextManager := context.NewContextManager(logger)
 	monitorMonitor := monitor.NewMonitorFromBootstrapConfig(bootstrap, logger)
 	dataData, cleanup, err := data.NewData(confData, logger, vectorService, documentProcessor, knowledgeSearcher, contextManager, monitorMonitor)
 	if err != nil {
@@ -44,16 +43,10 @@ func wireApp(confServer *conf.Server, confData *conf.Data, bootstrap *conf.Boots
 	}
 	customerRepo := repoimpl.NewCustomerRepo(dataData, logger)
 	customerUseCase := biz.NewCustomerUseCase(customerRepo, logger, knowledgeSearcher, documentProcessor, contextManager, monitorMonitor)
-    aiService, err := ai.NewAIServiceFromConfig(bootstrap, logger, knowledgeSearcher, contextManager, monitorMonitor)
-	if err != nil {
-		cleanup()
-		return nil, nil, err
-	}
 	sessionManager := session.NewMemorySessionManager(logger)
-	customerService := service.NewCustomerService(customerUseCase, aiService, sessionManager, monitorMonitor, logger)
+	customerService := service.NewCustomerService(customerUseCase, sessionManager, monitorMonitor, logger)
-	grpcServer := server.NewGRPCServer(confServer, customerService, logger)
 	httpServer := server.NewHTTPServer(confServer, customerService, logger)
-	app := newApp(logger, grpcServer, httpServer)
+	app := newApp(logger, httpServer)
 	return app, func() {
 		cleanup()
 	}, nil

eino-project/go.mod

@@ -46,6 +46,7 @@ require (
 	github.com/google/uuid v1.6.0 // indirect
 	github.com/goph/emperror v0.17.2 // indirect
 	github.com/gorilla/mux v1.8.1 // indirect
+	github.com/gorilla/websocket v1.5.3 // indirect
 	github.com/invopop/yaml v0.1.0 // indirect
 	github.com/josharian/intern v1.0.0 // indirect
 	github.com/json-iterator/go v1.1.12 // indirect

eino-project/go.sum

@@ -101,6 +101,7 @@ github.com/google/go-cmp v0.7.0/go.mod h1:pXiqmnSA92OHEEa9HXL2W4E7lf9JzCmGVUdgjX
 github.com/google/gofuzz v1.0.0/go.mod h1:dBl0BpW6vV/+mYPU4Po3pmUjxk6FQPldtuIdl/M65Eg=
 github.com/google/pprof v0.0.0-20210407192527-94a9f03dee38 h1:yAJXTCF9TqKcTiHJAE8dj7HMvPfh66eeA2JYW7eFpSE=
 github.com/google/pprof v0.0.0-20210407192527-94a9f03dee38/go.mod h1:kpwsk12EmLew5upagYY7GY0pfYCcupk39gWOCRROcvE=
+github.com/google/subcommands v1.2.0 h1:vWQspBTo2nEqTUFita5/KeEWlUL8kQObDFbub/EN9oE=
 github.com/google/subcommands v1.2.0/go.mod h1:ZjhPrFU+Olkh9WazFPsl27BQ4UPiG37m3yTrtFlrHVk=
 github.com/google/uuid v1.6.0 h1:NIvaJDMOsjHA8n1jAhLSgzrAzy1Hgr+hNrb57e+94F0=
 github.com/google/uuid v1.6.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
@@ -113,6 +114,8 @@ github.com/gopherjs/gopherjs v1.17.2/go.mod h1:pRRIvn/QzFLrKfvEz3qUuEhtE/zLCWfre
 github.com/gorilla/mux v1.8.0/go.mod h1:DVbg23sWSpFRCP0SfiEN6jmj59UnW/n46BH5rLB71So=
 github.com/gorilla/mux v1.8.1 h1:TuBL49tXwgrFYWhqrNgrUNEY92u81SPhu7sTdzQEiWY=
 github.com/gorilla/mux v1.8.1/go.mod h1:AKf9I4AEqPTmMytcMc0KkNouC66V3BtZ4qD5fmWSiMQ=
+github.com/gorilla/websocket v1.5.3 h1:saDtZ6Pbx/0u+bgYQ3q96pZgCzfhKXGPqt7kZ72aNNg=
+github.com/gorilla/websocket v1.5.3/go.mod h1:YR8l580nyteQvAITg2hZ9XVh4b55+EU/adAjf1fMHhE=
 github.com/hpcloud/tail v1.0.0/go.mod h1:ab1qPbhIpdTxEkNHXyeSf5vhxWSCs/tWer42PpOxQnU=
 github.com/invopop/yaml v0.1.0 h1:YW3WGUoJEXYfzWBjn00zIlrw7brGVD0fUKRYDPAPhrc=
 github.com/invopop/yaml v0.1.0/go.mod h1:2XuRLgs/ouIrW3XNzuNj7J3Nvu/Dig5MXvbCEdiBN3Q=
@@ -251,6 +254,8 @@ golang.org/x/mod v0.6.0-dev.0.20220419223038-86c51ed26bb4/go.mod h1:jJ57K6gSWd91
 golang.org/x/mod v0.8.0/go.mod h1:iBbtSCu2XBx23ZKBPSOrRkjjQPZFPuis4dIYUhu/chs=
 golang.org/x/mod v0.12.0/go.mod h1:iBbtSCu2XBx23ZKBPSOrRkjjQPZFPuis4dIYUhu/chs=
 golang.org/x/mod v0.14.0/go.mod h1:hTbmBsO62+eylJbnUtE2MGJUyE7QWk4xUqPFrRgJ+7c=
+golang.org/x/mod v0.25.0 h1:n7a+ZbQKQA/Ysbyb0/6IbB1H/X41mKgbhfv7AfG/44w=
+golang.org/x/mod v0.25.0/go.mod h1:IXM97Txy2VM4PJ3gI61r1YEk/gAj6zAHN3AdZt6S9Ww=
 golang.org/x/net v0.0.0-20180906233101-161cd47e91fd/go.mod h1:mL1N/T3taQHkDXs73rZJwtUhF3w3ftmwwsq0BUmARs4=
 golang.org/x/net v0.0.0-20190620200207-3b0461eec859/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s=
 golang.org/x/net v0.0.0-20210226172049-e18ecbb05110/go.mod h1:m0MpNAwzfU5UDzcl9v0D8zg8gWTRqZa9RBIspLL5mdg=

eino-project/internal/ai/ai.go

@@ -1,329 +0,0 @@
-package ai
-
-import (
-	"context"
-	"fmt"
-	"strings"
-	"time"
-
-	contextpkg "eino-project/internal/context"
-	"eino-project/internal/monitor"
-	"eino-project/internal/vector"
-
-	"github.com/cloudwego/eino/components/model"
-	"github.com/cloudwego/eino/schema"
-	"github.com/go-kratos/kratos/v2/log"
-)
-
-// AIService AI服务接口
-type AIService interface {
-	ProcessChat(ctx context.Context, message string, sessionID string) (string, error)
-	StreamChat(ctx context.Context, message string, sessionID string) (<-chan string, error)
-	// AnalyzeIntent 使用意图模型（qwen3:8b）做意图识别，仅返回固定标签之一
-	AnalyzeIntent(ctx context.Context, message string) (string, error)
-}
-
-// OllamaRequest Ollama API 请求结构
-type OllamaRequest struct {
-	Model  string `json:"model"`
-	Prompt string `json:"prompt"`
-	Stream bool   `json:"stream"`
-}
-
-// OllamaResponse Ollama API 响应结构
-type OllamaResponse struct {
-	Model     string `json:"model"`
-	Response  string `json:"response"`
-	Done      bool   `json:"done"`
-	CreatedAt string `json:"created_at"`
-}
-
-// aiService AI服务实现
-type aiService struct {
-    logger            log.Logger
-    chatModel         model.BaseChatModel
-    intentModel       model.BaseChatModel
-    knowledgeSearcher vector.KnowledgeSearcher
-    contextManager    contextpkg.ContextManager
-    monitor           interface {
-        RecordLLMUsage(ctx context.Context, usage *monitor.LLMUsage) error
-    }
-    chatModelName string
-}
-
-// NewAIService 创建AI服务实例
-func NewAIService(logger log.Logger, chatModel, intentModel model.BaseChatModel, knowledgeSearcher vector.KnowledgeSearcher, contextManager contextpkg.ContextManager) AIService {
-    return &aiService{
-        logger:            logger,
-        chatModel:         chatModel,
-        intentModel:       intentModel,
-        knowledgeSearcher: knowledgeSearcher,
-        contextManager:    contextManager,
-    }
-}
-
-// ProcessChat 处理聊天消息
-func (s *aiService) ProcessChat(ctx context.Context, message string, sessionID string) (string, error) {
-	log.Context(ctx).Infof("Processing chat message: %s for session: %s", message, sessionID)
-
-	// 1. 添加消息到上下文管理器
-	if s.contextManager != nil {
-		msg := contextpkg.Message{
-			Role:      "user",
-			Content:   message,
-			Timestamp: time.Now(),
-		}
-		s.contextManager.AddMessage(ctx, sessionID, msg)
-	}
-
-	// 2. 搜索相关知识库内容
-	var knowledgeContext string
-	if s.knowledgeSearcher != nil {
-		knowledgeResults, err := s.knowledgeSearcher.SearchKnowledge(ctx, message, 3)
-		if err == nil && len(knowledgeResults) > 0 {
-			var contextParts []string
-			for _, result := range knowledgeResults {
-				contextParts = append(contextParts, fmt.Sprintf("相关知识: %s", result.Document.Content))
-			}
-			knowledgeContext = strings.Join(contextParts, "\n")
-		}
-	}
-
-	// 3. 构建增强的聊天消息
-	enhancedMessage := message
-	if knowledgeContext != "" {
-		enhancedMessage = fmt.Sprintf("基于以下知识库内容回答用户问题:\n%s\n\n用户问题: %s", knowledgeContext, message)
-	}
-
-	messages := []*schema.Message{
-		{
-			Role:    schema.User,
-			Content: enhancedMessage,
-		},
-	}
-
-	// 4. 调用 Eino 聊天模型
-	start := time.Now()
-	response, err := s.chatModel.Generate(ctx, messages)
-	if err != nil {
-		log.Context(ctx).Warnf("Eino chat model call failed: %v, falling back to mock response", err)
-		// 如果 Eino 调用失败，返回模拟响应
-		return s.generateMockResponse(message), nil
-	}
-
-	if response == nil || response.Content == "" {
-		log.Context(ctx).Warn("Empty response from Eino chat model, falling back to mock response")
-		return s.generateMockResponse(message), nil
-	}
-
-	// 轻量化上报LLM使用：模型、token估算、延迟、知识命中等
-	if s.monitor != nil {
-		// eino 不直接暴露 tokens，这里用近似估算：中文字数/2，英文按空格词数*1
-		promptTokens := estimateTokens(enhancedMessage)
-		completionTokens := estimateTokens(response.Content)
-		// 使用专用意图模型识别意图（不使用自定义规则）
-		detectedIntent := "general_inquiry"
-		if intent, err := s.AnalyzeIntent(ctx, message); err == nil && intent != "" {
-			detectedIntent = intent
-		}
-		usage := &monitor.LLMUsage{
-			Model:            s.chatModelName,
-			SessionID:        sessionID,
-			UserID:           "default_user",
-			PromptPreview:    preview(enhancedMessage, 200),
-			PromptTokens:     promptTokens,
-			CompletionTokens: completionTokens,
-			TotalTokens:      promptTokens + completionTokens,
-			LatencyMS:        time.Since(start).Milliseconds(),
-			AgentThought:     "intent=" + detectedIntent,
-			KnowledgeHits:    countLines(knowledgeContext),
-			Metadata:         map[string]string{"source": "ai.ProcessChat"},
-			Timestamp:        time.Now(),
-		}
-		_ = s.monitor.RecordLLMUsage(ctx, usage)
-	}
-
-	return response.Content, nil
-}
-
-// StreamChat 流式处理聊天消息
-func (s *aiService) StreamChat(ctx context.Context, message string, sessionID string) (<-chan string, error) {
-	log.Context(ctx).Infof("Processing stream chat message: %s for session: %s", message, sessionID)
-
-	// 构建聊天消息
-	messages := []*schema.Message{
-		{
-			Role:    schema.User,
-			Content: message,
-		},
-	}
-
-	// 调用 Eino 流式聊天模型
-	start := time.Now()
-	streamReader, err := s.chatModel.Stream(ctx, messages)
-	if err != nil {
-		log.Context(ctx).Warnf("Eino stream chat model call failed: %v, falling back to mock stream response", err)
-		// 如果 Eino 流式调用失败，返回模拟流式响应
-		return s.generateMockStreamResponse(ctx, message), nil
-	}
-
-	// 创建响应通道
-	responseChan := make(chan string, 10)
-
-	// 启动 goroutine 处理流式响应
-	go func() {
-		defer close(responseChan)
-		defer streamReader.Close()
-
-		for {
-			chunk, err := streamReader.Recv()
-			if err != nil {
-				// 流结束或出错
-				return
-			}
-			if chunk != nil && chunk.Content != "" {
-				select {
-				case responseChan <- chunk.Content:
-				case <-ctx.Done():
-					return
-				}
-			}
-		}
-
-		// 汇总轻量化上报：以流结束后估算总tokens与延迟
-		if s.monitor != nil {
-			// 注意：为了避免阻塞，这里不去累计所有chunk；仅使用输入估算 + 总延迟
-			detectedIntent := "general_inquiry"
-			if intent, err := s.AnalyzeIntent(ctx, message); err == nil && intent != "" {
-				detectedIntent = intent
-			}
-			usage := &monitor.LLMUsage{
-				Model:         s.chatModelName,
-				SessionID:     sessionID,
-				UserID:        "default_user",
-				PromptPreview: preview(message, 200),
-				PromptTokens:  estimateTokens(message),
-				// completionTokens 无法准确统计，留空或基于时间估算可选
-				TotalTokens:  estimateTokens(message),
-				LatencyMS:    time.Since(start).Milliseconds(),
-				AgentThought: "intent=" + detectedIntent,
-				Metadata:     map[string]string{"source": "ai.StreamChat"},
-				Timestamp:    time.Now(),
-			}
-			_ = s.monitor.RecordLLMUsage(ctx, usage)
-		}
-	}()
-
-	return responseChan, nil
-}
-
-// estimateTokens 简易估算 tokens 数（不准确，但用于监控趋势足够）
-func estimateTokens(text string) int {
-	if text == "" {
-		return 0
-	}
-	// 粗略策略：英文按空格分词；中文按 rune 数/2
-	ascii := true
-	for _, r := range text {
-		if r > 127 {
-			ascii = false
-			break
-		}
-	}
-	if ascii {
-		// 英文：词数近似为token数
-		return len(strings.Fields(text))
-	}
-	// 中文：每2个汉字约1个token
-	return len([]rune(text)) / 2
-}
-
-func preview(text string, max int) string {
-	if len(text) <= max {
-		return text
-	}
-	return text[:max]
-}
-
-func countLines(text string) int {
-	if text == "" {
-		return 0
-	}
-	return len(strings.Split(text, "\n"))
-}
-
-// AnalyzeIntent 使用意图模型（qwen3:8b）进行意图识别，输出以下标签之一：
-// order_inquiry、product_inquiry、technical_support、general_inquiry
-func (s *aiService) AnalyzeIntent(ctx context.Context, message string) (string, error) {
-	if s.intentModel == nil {
-		return "general_inquiry", fmt.Errorf("intent model not configured")
-	}
-	instruction := "你是一个意图分类器。仅从以下标签中选择并输出一个，且只输出标签本身：order_inquiry、product_inquiry、technical_support、general_inquiry。用户消息：" + message
-	messages := []*schema.Message{
-		{Role: schema.User, Content: instruction},
-	}
-	resp, err := s.intentModel.Generate(ctx, messages)
-	if err != nil || resp == nil || resp.Content == "" {
-		return "general_inquiry", err
-	}
-	intent := strings.TrimSpace(strings.ToLower(resp.Content))
-	switch intent {
-	case "order_inquiry", "product_inquiry", "technical_support", "general_inquiry":
-		return intent, nil
-	}
-	// 输出不在预期集合，回退为general
-	return "general_inquiry", nil
-}
-
-// generateMockStreamResponse 生成模拟流式响应
-func (s *aiService) generateMockStreamResponse(ctx context.Context, message string) <-chan string {
-	responseChan := make(chan string, 10)
-
-	go func() {
-		defer close(responseChan)
-
-		// 获取模拟响应
-		mockResponse := s.generateMockResponse(message)
-
-		// 将响应分割成小块，模拟流式输出
-		words := strings.Fields(mockResponse)
-		for i, word := range words {
-			select {
-			case responseChan <- word:
-				if i < len(words)-1 {
-					// 在单词之间添加空格，除了最后一个单词
-					select {
-					case responseChan <- " ":
-					case <-ctx.Done():
-						return
-					}
-				}
-				// 模拟网络延迟
-				time.Sleep(50 * time.Millisecond)
-			case <-ctx.Done():
-				return
-			}
-		}
-	}()
-
-	return responseChan
-}
-
-// generateMockResponse 生成模拟响应
-func (s *aiService) generateMockResponse(message string) string {
-	// 简单的模拟响应逻辑
-	message = strings.ToLower(strings.TrimSpace(message))
-
-	switch {
-	case strings.Contains(message, "你好") || strings.Contains(message, "hello"):
-		return "你好！有什么我可以帮你的吗？😊"
-	case strings.Contains(message, "天气"):
-		return "抱歉，我无法获取实时天气信息。建议您查看天气预报应用。"
-	case strings.Contains(message, "时间"):
-		return fmt.Sprintf("当前时间是 %s", time.Now().Format("2006-01-02 15:04:05"))
-	case strings.Contains(message, "介绍"):
-		return "我是一个AI助手，可以帮助您回答问题、提供信息和进行对话。有什么我可以为您做的吗？"
-	default:
-		return "感谢您的消息！我正在学习中，暂时无法完全理解您的问题。请尝试用不同的方式表达，或者问一些简单的问题。"
-	}
-}

eino-project/internal/ai/provider.go

@@ -1,72 +0,0 @@
-package ai
-
-import (
-	"context"
-	"fmt"
-	"time"
-
-	"eino-project/internal/conf"
-	contextpkg "eino-project/internal/context"
-	"eino-project/internal/monitor"
-	"eino-project/internal/vector"
-
-	"github.com/cloudwego/eino-ext/components/model/ollama"
-	"github.com/go-kratos/kratos/v2/log"
-	"github.com/google/wire"
-)
-
-// ProviderSet is ai providers.
-var ProviderSet = wire.NewSet(NewAIServiceFromConfig)
-
-// NewAIServiceFromConfig 从配置创建AI服务
-func NewAIServiceFromConfig(c *conf.Bootstrap, logger log.Logger, knowledgeSearcher vector.KnowledgeSearcher, contextManager contextpkg.ContextManager, mon monitor.Monitor) (AIService, error) {
-	if c.Ai == nil || c.Ai.Ollama == nil {
-		return nil, fmt.Errorf("AI configuration is missing")
-	}
-
-	// 获取超时配置，默认60秒
-	timeout := 60 * time.Second
-	if c.Ai.Ollama.Timeout != nil {
-		timeout = c.Ai.Ollama.Timeout.AsDuration()
-	}
-
-	// 设置默认模型
-	chatModelName := "deepseek-v3.1:671b-cloud"
-	intentModelName := "qwen3:8b"
-
-	// 从配置中获取模型
-	if len(c.Ai.Ollama.Models) > 0 && c.Ai.Ollama.Models[0] != "" {
-		chatModelName = c.Ai.Ollama.Models[0]
-	}
-	if len(c.Ai.Ollama.Models) > 1 && c.Ai.Ollama.Models[1] != "" {
-		intentModelName = c.Ai.Ollama.Models[1]
-	}
-
-    // 创建聊天模型（满足 BaseChatModel 接口）
-    chatModel, err := ollama.NewChatModel(context.Background(), &ollama.ChatModelConfig{
-        BaseURL: c.Ai.Ollama.Endpoint,
-        Timeout: timeout,
-        Model:   chatModelName,
-    })
-	if err != nil {
-		return nil, fmt.Errorf("failed to create chat model: %w", err)
-	}
-
-    // 创建意图识别模型（满足 BaseChatModel 接口）
-    intentModel, err := ollama.NewChatModel(context.Background(), &ollama.ChatModelConfig{
-        BaseURL: c.Ai.Ollama.Endpoint,
-        Timeout: timeout,
-        Model:   intentModelName,
-    })
-	if err != nil {
-		return nil, fmt.Errorf("failed to create intent model: %w", err)
-	}
-
-	svc := NewAIService(logger, chatModel, intentModel, knowledgeSearcher, contextManager)
-	// 注入轻量监控，用于记录 LLM 使用
-	if m, ok := svc.(*aiService); ok {
-		m.monitor = mon
-		m.chatModelName = chatModelName
-	}
-	return svc, nil
-}

eino-project/internal/biz/customer.go

@@ -6,15 +6,13 @@ import (
 	"strings"
 	"time"
 
-	"eino-project/internal/vector"
+	contextpkg "eino-project/internal/domain/context"
-	contextpkg "eino-project/internal/context"
+	"eino-project/internal/domain/monitor"
-	"eino-project/internal/monitor"
+	"eino-project/internal/domain/vector"
 
 	"github.com/go-kratos/kratos/v2/log"
 )
 
-
-
 // CustomerRepo 智能客服数据仓库接口
 type CustomerRepo interface {
 	CheckSystemHealth(ctx context.Context) map[string]ServiceStatus
@@ -91,8 +89,6 @@ type OrderDetails struct {
 	NeedAI     bool      `json:"need_ai"`
 }
 
-
-
 // GetSystemStatus 获取系统状态
 func (uc *CustomerUseCase) GetSystemStatus(ctx context.Context) (*SystemStatus, error) {
 	uc.log.WithContext(ctx).Info("Getting system status")
@@ -311,8 +307,8 @@ func (uc *CustomerUseCase) GetKnowledgeSummary(ctx context.Context, query string
 // 保留函数以避免外部调用崩溃，但将其标记为 deprecated，并直接返回 general_inquiry
 // Deprecated: use AIService.AnalyzeIntent instead.
 func (uc *CustomerUseCase) AnalyzeUserIntent(ctx context.Context, message string) (string, error) {
-    uc.log.WithContext(ctx).Warn("AnalyzeUserIntent (rule-based) is deprecated, use AIService.AnalyzeIntent (LLM) instead")
+	uc.log.WithContext(ctx).Warn("AnalyzeUserIntent (rule-based) is deprecated, use AIService.AnalyzeIntent (LLM) instead")
-    return "general_inquiry", nil
+	return "general_inquiry", nil
 }
 
 // ProcessIntelligentChat 处理智能聊天（结合知识库）

eino-project/internal/conf/conf.pb.go

@@ -919,11 +919,8 @@ type Monitoring_CozeLoop struct {
 	Endpoint        string                 `protobuf:"bytes,1,opt,name=endpoint,proto3" json:"endpoint,omitempty"`
 	Enable          bool                   `protobuf:"varint,2,opt,name=enable,proto3" json:"enable,omitempty"`
 	MetricsInterval *durationpb.Duration   `protobuf:"bytes,3,opt,name=metrics_interval,json=metricsInterval,proto3" json:"metrics_interval,omitempty"`
-	// 新增：支持在配置文件中声明 Workspace 与 PAT（优先级高于环境变量）
+	unknownFields   protoimpl.UnknownFields
-	WorkspaceId   string `protobuf:"bytes,4,opt,name=workspace_id,json=workspaceId,proto3" json:"workspace_id,omitempty"`
+	sizeCache       protoimpl.SizeCache
-	ApiToken      string `protobuf:"bytes,5,opt,name=api_token,json=apiToken,proto3" json:"api_token,omitempty"`
-	unknownFields protoimpl.UnknownFields
-	sizeCache     protoimpl.SizeCache
 }
 
 func (x *Monitoring_CozeLoop) Reset() {
@@ -977,20 +974,6 @@ func (x *Monitoring_CozeLoop) GetMetricsInterval() *durationpb.Duration {
 	return nil
 }
 
-func (x *Monitoring_CozeLoop) GetWorkspaceId() string {
-	if x != nil {
-		return x.WorkspaceId
-	}
-	return ""
-}
-
-func (x *Monitoring_CozeLoop) GetApiToken() string {
-	if x != nil {
-		return x.ApiToken
-	}
-	return ""
-}
-
 var File_conf_conf_proto protoreflect.FileDescriptor
 
 const file_conf_conf_proto_rawDesc = "" +
@@ -1049,16 +1032,14 @@ const file_conf_conf_proto_rawDesc = "" +
 	"\n" +
 	"collection\x18\x02 \x01(\tR\n" +
 	"collection\x123\n" +
-	"\atimeout\x18\x03 \x01(\v2\x19.google.protobuf.DurationR\atimeout\"\x91\x02\n" +
+	"\atimeout\x18\x03 \x01(\v2\x19.google.protobuf.DurationR\atimeout\"\xd1\x01\n" +
 	"\n" +
 	"Monitoring\x12<\n" +
-	"\tcoze_loop\x18\x01 \x01(\v2\x1f.kratos.api.Monitoring.CozeLoopR\bcozeLoop\x1a\xc4\x01\n" +
+	"\tcoze_loop\x18\x01 \x01(\v2\x1f.kratos.api.Monitoring.CozeLoopR\bcozeLoop\x1a\x84\x01\n" +
 	"\bCozeLoop\x12\x1a\n" +
 	"\bendpoint\x18\x01 \x01(\tR\bendpoint\x12\x16\n" +
 	"\x06enable\x18\x02 \x01(\bR\x06enable\x12D\n" +
-	"\x10metrics_interval\x18\x03 \x01(\v2\x19.google.protobuf.DurationR\x0fmetricsInterval\x12!\n" +
+	"\x10metrics_interval\x18\x03 \x01(\v2\x19.google.protobuf.DurationR\x0fmetricsInterval\"K\n" +
-	"\fworkspace_id\x18\x04 \x01(\tR\vworkspaceId\x12\x1b\n" +
-	"\tapi_token\x18\x05 \x01(\tR\bapiToken\"K\n" +
 	"\x03Log\x12\x14\n" +
 	"\x05level\x18\x01 \x01(\tR\x05level\x12\x16\n" +
 	"\x06format\x18\x02 \x01(\tR\x06format\x12\x16\n" +

eino-project/internal/context/provider.go

@@ -1,16 +0,0 @@
-package context
-
-import (
-	"eino-project/internal/conf"
-
-	"github.com/go-kratos/kratos/v2/log"
-	"github.com/google/wire"
-)
-
-// ProviderSet is context providers.
-var ProviderSet = wire.NewSet(NewContextManagerFromBootstrapConfig)
-
-// NewContextManagerFromBootstrapConfig 从 Bootstrap 配置创建上下文管理器
-func NewContextManagerFromBootstrapConfig(c *conf.Bootstrap, logger log.Logger) ContextManager {
-	return NewContextManager(logger)
-}

eino-project/internal/data/data.go

@@ -6,9 +6,9 @@ import (
 	"time"
 
 	"eino-project/internal/conf"
-	contextpkg "eino-project/internal/context"
+	contextpkg "eino-project/internal/domain/context"
-	"eino-project/internal/monitor"
+	"eino-project/internal/domain/monitor"
-	"eino-project/internal/vector"
+	"eino-project/internal/domain/vector"
 
 	"github.com/go-kratos/kratos/v2/log"
 	"github.com/redis/go-redis/v9"
@@ -22,14 +22,14 @@ var ProviderSet = wire.NewSet(NewData)
 
 // Data 数据层结构
 type Data struct {
-	DB               *sql.DB
+	DB                *sql.DB
-	RDB              *redis.Client
+	RDB               *redis.Client
-	VectorService    vector.VectorService
+	VectorService     vector.VectorService
-	DocProcessor     vector.DocumentProcessor
+	DocProcessor      vector.DocumentProcessor
 	KnowledgeSearcher vector.KnowledgeSearcher
-	log              *log.Helper
+	log               *log.Helper
-	ContextManager   contextpkg.ContextManager
+	ContextManager    contextpkg.ContextManager
-	Monitor          monitor.Monitor
+	Monitor           monitor.Monitor
 }
 
 // NewData 创建数据层
@@ -67,14 +67,14 @@ func NewData(c *conf.Data, logger log.Logger, vs vector.VectorService, dp vector
 	}
 
 	d := &Data{
-		DB:               db,
+		DB:                db,
-		RDB:              rdb,
+		RDB:               rdb,
-		VectorService:    vs,
+		VectorService:     vs,
-		DocProcessor:     dp,
+		DocProcessor:      dp,
 		KnowledgeSearcher: ks,
-		log:              helper,
+		log:               helper,
-		ContextManager:   contextManager,
+		ContextManager:    contextManager,
-		Monitor:          monitor,
+		Monitor:           monitor,
 	}
 
 	cleanup := func() {

eino-project/internal/domain/agent/intent/intent.go

@@ -0,0 +1,17 @@
+package intent
+
+import (
+    "context"
+)
+
+type IntentAgent interface {
+    Classify(ctx context.Context, message string) (string, error)
+}
+
+type passthroughAgent struct{}
+
+func NewPassthrough() IntentAgent { return &passthroughAgent{} }
+
+func (p *passthroughAgent) Classify(ctx context.Context, message string) (string, error) {
+    return "general_inquiry", nil
+}

eino-project/internal/domain/agent/product/handler.go

@@ -0,0 +1,63 @@
+package product
+
+import (
+    "encoding/json"
+    "net/http"
+    "strings"
+
+    tool "eino-project/internal/tools/product"
+)
+
+type Handler struct {
+    tool tool.Tool
+}
+
+func NewHandler(t tool.Tool) *Handler { return &Handler{tool: t} }
+
+type queryRequest struct {
+    Query struct {
+        ID   string `json:"id"`
+        Name string `json:"name"`
+    } `json:"query"`
+    Page int `json:"page,omitempty"`
+    Size int `json:"size,omitempty"`
+}
+
+type queryResponse struct {
+    Items  []*tool.Product `json:"items"`
+    Source string          `json:"source"`
+    Count  int             `json:"count"`
+}
+
+func (h *Handler) ServeHTTP(w http.ResponseWriter, r *http.Request) {
+    if r.Method != http.MethodPost {
+        http.Error(w, "Method not allowed", http.StatusMethodNotAllowed)
+        return
+    }
+
+    var req queryRequest
+    if err := json.NewDecoder(r.Body).Decode(&req); err != nil {
+        http.Error(w, "Invalid request body", http.StatusBadRequest)
+        return
+    }
+
+    var items []*tool.Product
+    src := ""
+
+    if id := strings.TrimSpace(req.Query.ID); id != "" {
+        if p, _ := h.tool.GetByID(r.Context(), id); p != nil {
+            items = append(items, p)
+        }
+        src = "id"
+    } else if name := strings.TrimSpace(req.Query.Name); name != "" {
+        res, _ := h.tool.SearchByName(r.Context(), name)
+        items = res
+        src = "name"
+    } else {
+        http.Error(w, "query.id or query.name required", http.StatusBadRequest)
+        return
+    }
+
+    w.Header().Set("Content-Type", "application/json")
+    json.NewEncoder(w).Encode(&queryResponse{Items: items, Source: src, Count: len(items)})
+}

eino-project/internal/domain/llm/llm.go

@@ -0,0 +1,44 @@
+package llm
+
+import (
+	"eino-project/internal/conf"
+	"sync"
+
+	"github.com/cloudwego/eino/components/model"
+)
+
+type LLM interface {
+	Chat() (model.BaseChatModel, error)
+	Intent() (model.BaseChatModel, error)
+}
+
+type llm struct {
+	cfg        *conf.Bootstrap
+	onceChat   sync.Once
+	onceIntent sync.Once
+	chat       model.BaseChatModel
+	intent     model.BaseChatModel
+	cache      map[string]model.BaseChatModel
+}
+
+func NewLLM(cfg *conf.Bootstrap) LLM {
+	return &llm{cfg: cfg, cache: make(map[string]model.BaseChatModel)}
+}
+
+// 获取Ollama聊天模型实例
+func (r *llm) Chat() (model.BaseChatModel, error) {
+	var err error
+	r.onceChat.Do(func() {
+		r.chat, err = newOllamaChatModel(r.cfg)
+	})
+	return r.chat, err
+}
+
+// 获取Ollama意图识别模型实例
+func (r *llm) Intent() (model.BaseChatModel, error) {
+	var err error
+	r.onceIntent.Do(func() {
+		r.intent, err = newOllamaIntentModel(r.cfg)
+	})
+	return r.intent, err
+}

eino-project/internal/domain/llm/ollama_chat.go

@@ -0,0 +1,37 @@
+package llm
+
+import (
+	"context"
+	"fmt"
+	"time"
+
+	"eino-project/internal/conf"
+
+	"github.com/cloudwego/eino-ext/components/model/ollama"
+	"github.com/cloudwego/eino/components/model"
+)
+
+// newOllamaChatModel Ollama聊天模型实例
+func newOllamaChatModel(c *conf.Bootstrap) (model.BaseChatModel, error) {
+	if c == nil || c.Ai == nil || c.Ai.Ollama == nil {
+		return nil, fmt.Errorf("AI configuration is missing")
+	}
+	timeout := 60 * time.Second
+	if c.Ai.Ollama.Timeout != nil {
+		timeout = c.Ai.Ollama.Timeout.AsDuration()
+	}
+	modelName := "deepseek-v3.1:671b-cloud"
+	if len(c.Ai.Ollama.Models) > 0 && c.Ai.Ollama.Models[0] != "" {
+		modelName = c.Ai.Ollama.Models[0]
+	}
+	model, err := ollama.NewChatModel(context.Background(), &ollama.ChatModelConfig{
+		BaseURL: c.Ai.Ollama.Endpoint,
+		Timeout: timeout,
+		Model:   modelName,
+	})
+	if err != nil {
+		return nil, fmt.Errorf("failed to create chat model: %w", err)
+	}
+
+	return model, nil
+}

eino-project/internal/domain/llm/ollama_intent.go

@@ -0,0 +1,36 @@
+package llm
+
+import (
+	"context"
+	"fmt"
+	"time"
+
+	"eino-project/internal/conf"
+
+	"github.com/cloudwego/eino-ext/components/model/ollama"
+	"github.com/cloudwego/eino/components/model"
+)
+
+// newOllamaIntentModel Ollama意图识别模型实例
+func newOllamaIntentModel(c *conf.Bootstrap) (model.BaseChatModel, error) {
+	if c == nil || c.Ai == nil || c.Ai.Ollama == nil {
+		return nil, fmt.Errorf("AI configuration is missing")
+	}
+	timeout := 60 * time.Second
+	if c.Ai.Ollama.Timeout != nil {
+		timeout = c.Ai.Ollama.Timeout.AsDuration()
+	}
+	modelName := "qwen3:8b"
+	if len(c.Ai.Ollama.Models) > 1 && c.Ai.Ollama.Models[1] != "" {
+		modelName = c.Ai.Ollama.Models[1]
+	}
+	model, err := ollama.NewChatModel(context.Background(), &ollama.ChatModelConfig{
+		BaseURL: c.Ai.Ollama.Endpoint,
+		Timeout: timeout,
+		Model:   modelName,
+	})
+	if err != nil {
+		return nil, fmt.Errorf("failed to create intent model: %w", err)
+	}
+	return model, nil
+}

eino-project/internal/domain/monitor/coze_loop.go

@@ -1,95 +1,92 @@
 package monitor
 
 import (
-    "context"
+	"context"
-    "time"
+	"time"
 
-    "eino-project/internal/conf"
+	"eino-project/internal/conf"
 
-    cozeloop "github.com/coze-dev/cozeloop-go"
+	cozeloop "github.com/coze-dev/cozeloop-go"
-    "github.com/go-kratos/kratos/v2/log"
+	"github.com/go-kratos/kratos/v2/log"
 )
 
 // cozeLoopMonitor 是一个包装器：在调用本地监控逻辑的同时，将关键指标以 Trace 形式上报到 Coze Loop
 type cozeLoopMonitor struct {
-    base   Monitor
+	base   Monitor
-    client cozeloop.Client
+	client cozeloop.Client
-    log    *log.Helper
+	log    *log.Helper
 }
 
 // NewCozeLoopMonitor 根据配置初始化 Coze Loop 客户端并返回包装后的监控器
 func NewCozeLoopMonitor(base Monitor, cfg *conf.Monitoring_CozeLoop, logger log.Logger) Monitor {
-    helper := log.NewHelper(logger)
+	helper := log.NewHelper(logger)
 
-    ctx := context.Background()
+	ctx := context.Background()
 
-    // 按官方 simple demo 风格，直接从环境变量初始化客户端
+	// 按官方 simple demo 风格，直接从环境变量初始化客户端
-    client, err := cozeloop.NewClient()
+	client, err := cozeloop.NewClient()
-    if err != nil {
+	if err != nil {
-        helper.WithContext(ctx).Warnf("init CozeLoop client failed: %v, fallback to base monitor", err)
+		helper.WithContext(ctx).Warnf("init CozeLoop client failed: %v, fallback to base monitor", err)
-        return base
+		return base
-    }
+	}
 
-    m := &cozeLoopMonitor{
+	m := &cozeLoopMonitor{
-        base:   base,
+		base:   base,
-        client: client,
+		client: client,
-        log:    helper,
+		log:    helper,
-    }
+	}
 
 	return m
 }
 
-// loopReportMetrics 定期采集本地指标并以 Trace 形式上报
-// 已移除周期性指标上报逻辑，保留最小实现
-
 // RecordRequest 记录请求并上报简化 Trace
 func (m *cozeLoopMonitor) RecordRequest(ctx context.Context, requestType string, duration time.Duration, success bool) error {
-    if err := m.base.RecordRequest(ctx, requestType, duration, success); err != nil {
+	if err := m.base.RecordRequest(ctx, requestType, duration, success); err != nil {
-        return err
+		return err
-    }
+	}
-    // 简化：不生成trace
+	// 简化：不生成trace
-    return nil
+	return nil
 }
 
 // RecordAIRequest 记录 AI 请求并上报简化 Trace
 func (m *cozeLoopMonitor) RecordAIRequest(ctx context.Context, duration time.Duration, success bool) error {
-    if err := m.base.RecordAIRequest(ctx, duration, success); err != nil {
+	if err := m.base.RecordAIRequest(ctx, duration, success); err != nil {
-        return err
+		return err
-    }
+	}
-    // 简化：不生成trace
+	// 简化：不生成trace
-    return nil
+	return nil
 }
 
 // RecordVectorOperation 记录向量操作并上报简化 Trace
 func (m *cozeLoopMonitor) RecordVectorOperation(ctx context.Context, operation string, success bool) error {
-    if err := m.base.RecordVectorOperation(ctx, operation, success); err != nil {
+	if err := m.base.RecordVectorOperation(ctx, operation, success); err != nil {
-        return err
+		return err
-    }
+	}
-    // 简化：不生成trace
+	// 简化：不生成trace
-    return nil
+	return nil
 }
 
 // RecordSessionOperation 记录会话操作并上报简化 Trace
 func (m *cozeLoopMonitor) RecordSessionOperation(ctx context.Context, operation string) error {
-    if err := m.base.RecordSessionOperation(ctx, operation); err != nil {
+	if err := m.base.RecordSessionOperation(ctx, operation); err != nil {
-        return err
+		return err
-    }
+	}
-    // 简化：不生成trace
+	// 简化：不生成trace
-    return nil
+	return nil
 }
 
 // RecordKnowledgeOperation 记录知识库操作并上报简化 Trace
 func (m *cozeLoopMonitor) RecordKnowledgeOperation(ctx context.Context, operation string) error {
-    if err := m.base.RecordKnowledgeOperation(ctx, operation); err != nil {
+	if err := m.base.RecordKnowledgeOperation(ctx, operation); err != nil {
-        return err
+		return err
-    }
+	}
-    // 简化：不生成trace
+	// 简化：不生成trace
-    return nil
+	return nil
 }
 
 // 透传LLM使用事件（不生成trace，不做周期性上报）
 func (m *cozeLoopMonitor) RecordLLMUsage(ctx context.Context, usage *LLMUsage) error {
-    return m.base.RecordLLMUsage(ctx, usage)
+	return m.base.RecordLLMUsage(ctx, usage)
 }
 
 // GetMetrics 透传本地指标（不直接从 Coze Loop 读取）
@@ -99,10 +96,10 @@ func (m *cozeLoopMonitor) GetMetrics(ctx context.Context) (*Metrics, error) {
 
 // CreateAlert 创建告警（不生成trace，保持简化）
 func (m *cozeLoopMonitor) CreateAlert(ctx context.Context, alertType, level, title, message string, metadata map[string]interface{}) error {
-    if err := m.base.CreateAlert(ctx, alertType, level, title, message, metadata); err != nil {
+	if err := m.base.CreateAlert(ctx, alertType, level, title, message, metadata); err != nil {
-        return err
+		return err
-    }
+	}
-    return nil
+	return nil
 }
 
 // GetAlerts 透传本地告警

eino-project/internal/domain/provier_set.go

@@ -0,0 +1,27 @@
+package domain
+
+import (
+	"eino-project/internal/domain/context"
+	"eino-project/internal/domain/llm"
+	"eino-project/internal/domain/monitor"
+	"eino-project/internal/domain/session"
+	"eino-project/internal/domain/tools"
+	"eino-project/internal/domain/vector"
+	"eino-project/internal/domain/workflow"
+
+	"github.com/google/wire"
+)
+
+// ProviderSet is domain providers.
+var ProviderSet = wire.NewSet(
+	context.NewContextManager,
+	llm.NewLLM,
+	monitor.NewMonitorFromBootstrapConfig,
+	session.NewMemorySessionManager,
+	tools.NewMemoryProductTool,
+	workflow.NewChatWorkflow,
+
+	vector.NewVectorServiceFromBootstrapConfig,
+	vector.NewDocumentProcessor,
+	vector.NewKnowledgeSearcher,
+)

eino-project/internal/domain/tools/product.go

@@ -0,0 +1,54 @@
+package tools
+
+import (
+	"context"
+	"strings"
+)
+
+type Product struct {
+	ID          string  `json:"id"`
+	Name        string  `json:"name"`
+	Price       float64 `json:"price"`
+	Description string  `json:"description"`
+}
+
+type Tool interface {
+	GetByID(ctx context.Context, id string) (*Product, error)
+	SearchByName(ctx context.Context, name string) ([]*Product, error)
+}
+
+type memoryTool struct {
+	items []*Product
+}
+
+func NewMemoryProductTool() Tool {
+	return &memoryTool{items: []*Product{
+		{ID: "P001", Name: "手机 A1", Price: 1999, Description: "入门级智能手机"},
+		{ID: "P002", Name: "手机 Pro X", Price: 5999, Description: "旗舰级拍照手机"},
+		{ID: "P003", Name: "笔记本 M3", Price: 8999, Description: "轻薄高性能笔记本"},
+		{ID: "P004", Name: "耳机 Air", Price: 1299, Description: "降噪真无线耳机"},
+	}}
+}
+
+func (m *memoryTool) GetByID(ctx context.Context, id string) (*Product, error) {
+	for _, it := range m.items {
+		if it.ID == id {
+			return it, nil
+		}
+	}
+	return nil, nil
+}
+
+func (m *memoryTool) SearchByName(ctx context.Context, name string) ([]*Product, error) {
+	if name == "" {
+		return nil, nil
+	}
+	q := strings.ToLower(name)
+	var out []*Product
+	for _, it := range m.items {
+		if strings.Contains(strings.ToLower(it.Name), q) {
+			out = append(out, it)
+		}
+	}
+	return out, nil
+}

eino-project/internal/domain/workflow/chat_workflow.go

@@ -0,0 +1,104 @@
+package workflow
+
+import (
+	"context"
+	"fmt"
+	"strings"
+	"time"
+
+	contextpkg "eino-project/internal/domain/context"
+	"eino-project/internal/domain/llm"
+	"eino-project/internal/domain/vector"
+
+	"github.com/cloudwego/eino/schema"
+)
+
+type ChatWorkflow interface {
+	Chat(ctx context.Context, message string, sessionID string) (string, error)
+	Stream(ctx context.Context, message string, sessionID string) (<-chan string, error)
+}
+
+type chatWorkflow struct {
+	models   llm.LLM
+	searcher vector.KnowledgeSearcher
+	ctxMgr   contextpkg.ContextManager
+}
+
+func NewChatWorkflow(models llm.LLM, searcher vector.KnowledgeSearcher, ctxMgr contextpkg.ContextManager) ChatWorkflow {
+	return &chatWorkflow{models: models, searcher: searcher, ctxMgr: ctxMgr}
+}
+
+func (w *chatWorkflow) Chat(ctx context.Context, message string, sessionID string) (string, error) {
+	if w.ctxMgr != nil {
+		w.ctxMgr.AddMessage(ctx, sessionID, contextpkg.Message{Role: "user", Content: message, Timestamp: time.Now()})
+	}
+
+	var knowledgeContext string
+	if w.searcher != nil {
+		results, err := w.searcher.SearchKnowledge(ctx, message, 3)
+		if err == nil && len(results) > 0 {
+			var parts []string
+			for _, r := range results {
+				parts = append(parts, fmt.Sprintf("相关知识: %s", r.Document.Content))
+			}
+			knowledgeContext = strings.Join(parts, "\n")
+		}
+	}
+
+	enhanced := message
+	if knowledgeContext != "" {
+		enhanced = fmt.Sprintf("基于以下知识库内容回答用户问题:\n%s\n\n用户问题: %s", knowledgeContext, message)
+	}
+
+	msgs := []*schema.Message{{Role: schema.User, Content: enhanced}}
+	chatModel, err := w.models.Chat()
+	if err != nil {
+		return "", err
+	}
+	resp, err := chatModel.Generate(ctx, msgs)
+	if err != nil || resp == nil {
+		return "", err
+	}
+	return resp.Content, nil
+}
+
+func (w *chatWorkflow) Stream(ctx context.Context, message string, sessionID string) (<-chan string, error) {
+	var knowledgeContext string
+	if w.searcher != nil {
+		results, err := w.searcher.SearchKnowledge(ctx, message, 3)
+		if err == nil && len(results) > 0 {
+			var parts []string
+			for _, r := range results {
+				parts = append(parts, fmt.Sprintf("相关知识: %s", r.Document.Content))
+			}
+			knowledgeContext = strings.Join(parts, "\n")
+		}
+	}
+	enhanced := message
+	if knowledgeContext != "" {
+		enhanced = fmt.Sprintf("基于以下知识库内容回答用户问题:\n%s\n\n用户问题: %s", knowledgeContext, message)
+	}
+	msgs := []*schema.Message{{Role: schema.User, Content: enhanced}}
+	chatModel, err := w.models.Chat()
+	if err != nil {
+		return nil, err
+	}
+	reader, err := chatModel.Stream(ctx, msgs)
+	if err != nil {
+		return nil, err
+	}
+	ch := make(chan string, 8)
+	go func() {
+		defer close(ch)
+		for {
+			chunk, err := reader.Recv()
+			if err != nil {
+				return
+			}
+			if chunk != nil && chunk.Content != "" {
+				ch <- chunk.Content
+			}
+		}
+	}()
+	return ch, nil
+}

eino-project/internal/server/grpc.go

@@ -1,32 +0,0 @@
-package server
-
-import (
-	v1 "eino-project/api/customer/v1"
-	"eino-project/internal/conf"
-	"eino-project/internal/service"
-
-	"github.com/go-kratos/kratos/v2/log"
-	"github.com/go-kratos/kratos/v2/middleware/recovery"
-	"github.com/go-kratos/kratos/v2/transport/grpc"
-)
-
-// NewGRPCServer new a gRPC server.
-func NewGRPCServer(c *conf.Server, customerService *service.CustomerService, logger log.Logger) *grpc.Server {
-	var opts = []grpc.ServerOption{
-		grpc.Middleware(
-			recovery.Recovery(),
-		),
-	}
-	if c.Grpc.Network != "" {
-		opts = append(opts, grpc.Network(c.Grpc.Network))
-	}
-	if c.Grpc.Addr != "" {
-		opts = append(opts, grpc.Address(c.Grpc.Addr))
-	}
-	if c.Grpc.Timeout != nil {
-		opts = append(opts, grpc.Timeout(c.Grpc.Timeout.AsDuration()))
-	}
-	srv := grpc.NewServer(opts...)
-	v1.RegisterCustomerServiceServer(srv, customerService)
-	return srv
-}

eino-project/internal/server/http.go

@@ -4,6 +4,9 @@ import (
 	v1 "eino-project/api/customer/v1"
 	"eino-project/internal/conf"
 	"eino-project/internal/service"
+	nethttp "net/http"
+
+	"github.com/gorilla/websocket"
 
 	"github.com/go-kratos/kratos/v2/log"
 	"github.com/go-kratos/kratos/v2/middleware/recovery"
@@ -28,11 +31,27 @@ func NewHTTPServer(c *conf.Server, customerService *service.CustomerService, log
 	}
 	srv := http.NewServer(opts...)
 
-	// 注册标准的gRPC-Gateway路由
+	// 注册HTTP路由
 	v1.RegisterCustomerServiceHTTPServer(srv, customerService)
 
 	// 添加SSE流式聊天的自定义路由
 	srv.HandleFunc("/api/chat/stream", customerService.HandleStreamChat)
 
+	// 商品查询 Agent 路由
+	srv.HandleFunc("/api/agents/product/query", customerService.HandleProductQuery)
+
+	// WebSocket 聊天路由（/api/chat/ws）
+	srv.HandleFunc("/api/chat/ws", func(w nethttp.ResponseWriter, r *nethttp.Request) {
+		upgrader := websocket.Upgrader{CheckOrigin: func(r *nethttp.Request) bool { return true }}
+		conn, err := upgrader.Upgrade(w, r, nil)
+		if err != nil {
+			return
+		}
+		customerService.HandleWebSocketChat(conn)
+	})
+
+	// 订单诊断工作流（直接调用流水线输出）
+	srv.HandleFunc("/api/workflow/order/diagnosis", customerService.HandleOrderDiagnosis)
+
 	return srv
 }

eino-project/internal/server/server.go

@@ -1,8 +1,8 @@
 package server
 
 import (
-	"github.com/google/wire"
+    "github.com/google/wire"
 )
 
 // ProviderSet is server providers.
-var ProviderSet = wire.NewSet(NewGRPCServer, NewHTTPServer)
+var ProviderSet = wire.NewSet(NewHTTPServer)

eino-project/internal/service/customer.go

@@ -7,11 +7,13 @@ import (
 	"net/http"
 	"time"
 
+	"github.com/gorilla/websocket"
+
 	pb "eino-project/api/customer/v1"
-	"eino-project/internal/ai"
 	"eino-project/internal/biz"
-	"eino-project/internal/monitor"
+	"eino-project/internal/domain/monitor"
-	"eino-project/internal/session"
+	"eino-project/internal/domain/session"
+	wf "eino-project/internal/domain/workflow"
 
 	"github.com/go-kratos/kratos/v2/log"
 )
@@ -21,23 +23,24 @@ type CustomerService struct {
 	pb.UnimplementedCustomerServiceServer
 
 	customerUseCase *biz.CustomerUseCase
-	aiService       ai.AIService
 	sessionManager  session.SessionManager
 	monitor         monitor.Monitor
 	log             *log.Helper
+	chatWorkflow    wf.ChatWorkflow
 }
 
 // NewCustomerService 创建智能客服服务
-func NewCustomerService(customerUseCase *biz.CustomerUseCase, aiService ai.AIService, sessionManager session.SessionManager, monitor monitor.Monitor, logger log.Logger) *CustomerService {
+func NewCustomerService(customerUseCase *biz.CustomerUseCase, sessionManager session.SessionManager, monitor monitor.Monitor, logger log.Logger) *CustomerService {
 	return &CustomerService{
 		customerUseCase: customerUseCase,
-		aiService:       aiService,
 		sessionManager:  sessionManager,
 		monitor:         monitor,
 		log:             log.NewHelper(logger),
 	}
 }
 
+func (s *CustomerService) SetChatWorkflow(w wf.ChatWorkflow) { s.chatWorkflow = w }
+
 // SystemStatus 系统状态检查
 func (s *CustomerService) SystemStatus(ctx context.Context, req *pb.SystemStatusRequest) (*pb.SystemStatusResponse, error) {
 	s.log.WithContext(ctx).Info("SystemStatus called")
@@ -97,28 +100,16 @@ func (s *CustomerService) Chat(ctx context.Context, req *pb.ChatRequest) (*pb.Ch
 		}
 	}
 
-	// 分析用户意图（使用 qwen3:8b LLM，不使用自定义规则）
+	// 可在工作流内部进行意图识别
-	intent := "general_inquiry"
-	if aiIntent, err := s.aiService.AnalyzeIntent(ctx, req.Message); err == nil && aiIntent != "" {
-		intent = aiIntent
-	} else if err != nil {
-		s.log.Errorf("Failed to analyze user intent via LLM: %v", err)
-	}
 
 	var aiResponse string
 
-	// 根据意图选择处理方式
+	// 使用 ChatWorkflow 路由处理（意图判断 + 调用相应 Agent/Workflow）
-	if intent == "knowledge_inquiry" || intent == "technical_support" {
+	if s.chatWorkflow != nil {
-		// 使用智能问答（结合知识库）
+		aiResponse, err = s.chatWorkflow.Chat(ctx, req.Message, req.SessionId)
-		aiResponse, err = s.customerUseCase.ProcessIntelligentChat(ctx, req.Message, req.SessionId)
-		if err != nil {
-			s.log.Errorf("Intelligent chat error: %v", err)
-			// 降级到普通AI服务
-			aiResponse, err = s.aiService.ProcessChat(ctx, req.Message, req.SessionId)
-		}
 	} else {
-		// 使用普通AI服务
+		aiResponse = ""
-		aiResponse, err = s.aiService.ProcessChat(ctx, req.Message, req.SessionId)
+		err = fmt.Errorf("chat workflow not configured")
 	}
 
 	if err != nil {
@@ -301,9 +292,14 @@ func (s *CustomerService) StreamChat(req *pb.StreamChatRequest, stream pb.Custom
 		}
 	}
 
-	// 使用AI服务进行流式对话
+	// 使用工作流进行流式对话
 	var fullResponse string
-	responseChan, err := s.aiService.StreamChat(ctx, req.Message, req.SessionId)
+	var responseChan <-chan string
+	if s.chatWorkflow != nil {
+		responseChan, err = s.chatWorkflow.Stream(ctx, req.Message, req.SessionId)
+	} else {
+		err = fmt.Errorf("chat workflow not configured")
+	}
 	if err != nil {
 		s.log.Errorf("AI stream service error: %v", err)
 		return s.fallbackStreamResponse(req, stream)
@@ -445,9 +441,9 @@ func (s *CustomerService) HandleStreamChat(w http.ResponseWriter, r *http.Reques
 		}
 	}
 
-	// 使用AI服务进行流式对话
+	// 使用工作流进行流式对话
 	var fullResponse string
-	responseChan, err := s.aiService.StreamChat(ctx, req.Message, req.SessionID)
+	responseChan, err := s.chatWorkflow.Stream(ctx, req.Message, req.SessionID)
 	if err != nil {
 		s.log.Errorf("AI stream service error: %v", err)
 		// 降级到模拟响应
@@ -548,3 +544,68 @@ func (s *CustomerService) fallbackSSEResponse(w http.ResponseWriter, flusher htt
 	fmt.Fprintf(w, "data: {\"type\":\"end\",\"session_id\":\"%s\"}\n\n", sessionID)
 	flusher.Flush()
 }
+
+// HandleWebSocketChat WebSocket 聊天处理器
+func (s *CustomerService) HandleWebSocketChat(conn *websocket.Conn) {
+	defer conn.Close()
+	ctx := context.Background()
+	for {
+		mt, msg, err := conn.ReadMessage()
+		if err != nil {
+			return
+		}
+		if mt != websocket.TextMessage {
+			continue
+		}
+		var req struct {
+			Message   string `json:"message"`
+			SessionID string `json:"session_id"`
+		}
+		if json.Unmarshal(msg, &req) != nil {
+			continue
+		}
+		ch, err := s.chatWorkflow.Stream(ctx, req.Message, req.SessionID)
+		if err != nil {
+			_ = conn.WriteMessage(websocket.TextMessage, []byte(`{"type":"error"}`))
+			continue
+		}
+		for chunk := range ch {
+			resp := map[string]interface{}{"type": "chunk", "content": chunk, "session_id": req.SessionID}
+			b, _ := json.Marshal(resp)
+			_ = conn.WriteMessage(websocket.TextMessage, b)
+		}
+		_ = conn.WriteMessage(websocket.TextMessage, []byte(`{"type":"end"}`))
+	}
+}
+
+// HandleOrderDiagnosis 订单诊断工作流入口
+func (s *CustomerService) HandleOrderDiagnosis(w http.ResponseWriter, r *http.Request) {
+	if r.Method != http.MethodPost {
+		http.Error(w, "Method not allowed", http.StatusMethodNotAllowed)
+		return
+	}
+	var req struct {
+		OrderID   string `json:"order_id"`
+		Message   string `json:"message"`
+		SessionID string `json:"session_id"`
+	}
+	if err := json.NewDecoder(r.Body).Decode(&req); err != nil {
+		http.Error(w, "Invalid request body", http.StatusBadRequest)
+		return
+	}
+	ctx := r.Context()
+	// 简化：直接调用工作流的 Chat（可替换为订单专用流水线）
+	resp, err := s.chatWorkflow.Chat(ctx, fmt.Sprintf("订单诊断: %s\n%s", req.OrderID, req.Message), req.SessionID)
+	if err != nil {
+		http.Error(w, "Workflow error", http.StatusInternalServerError)
+		return
+	}
+	w.Header().Set("Content-Type", "application/json")
+	json.NewEncoder(w).Encode(map[string]interface{}{"order_id": req.OrderID, "result": resp})
+}
+
+// HandleProductQuery 商品查询处理
+func (s *CustomerService) HandleProductQuery(w http.ResponseWriter, r *http.Request) {
+
+	json.NewEncoder(w).Encode(map[string]interface{}{})
+}

eino-project/internal/session/provider.go

@@ -1,8 +0,0 @@
-package session
-
-import (
-	"github.com/google/wire"
-)
-
-// ProviderSet is session providers.
-var ProviderSet = wire.NewSet(NewMemorySessionManager)