add

langchain-project/.env

@@ -18,10 +18,15 @@ CHAT_MODEL_BASE_URL=http://localhost:11434
 CHAT_MODEL=deepseek-v3.1:671b-cloud
 
 # LangSmith监控配置
+
 LANGCHAIN_TRACING_V2=true
-LANGCHAIN_API_KEY=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
+LANGCHAIN_API_KEY=lsv2_pt_e6a2f0ca51874efcacfa0799146d698b_f09cd87b
 LANGCHAIN_PROJECT=default
 
+# LANGSMITH_TRACING=true
+# LANGSMITH_API_KEY="<your-langsmith-api-key>"
+# LANGSMITH_WORKSPACE_ID="<your-workspace-id>"
+
 # 业务系统API配置
 ORDER_API_BASE_URL=https://api.example.com
 

langchain-project/logs/langchain_bot_20251119.log

@@ -0,0 +1,46 @@
+2025-11-19 11:25:43,697 | INFO     | langsmith_service | info:107 | ✅ LangSmith 监控已启用
+2025-11-19 11:25:43,818 | INFO     | ollama_service | info:107 | ✅ Ollama模型初始化成功
+2025-11-19 11:25:43,819 | INFO     | ollama_service | info:107 | 路由模型: qwen3:8b
+2025-11-19 11:25:43,819 | INFO     | ollama_service | info:107 | 对话模型: deepseek-v3.1:671b-cloud
+2025-11-19 11:25:44,701 | INFO     | workflow | info:107 | 🔗 SimpleChatChain 初始化完成
+2025-11-19 11:25:44,708 | INFO     | workflow | info:107 | 📊 工作流图构建完成
+2025-11-19 11:25:44,708 | INFO     | workflow | info:107 | 🚀 ChatWorkflow 初始化完成
+2025-11-19 11:25:44,708 | INFO     | workflow | info:107 | 🤖 使用 LangGraph 工作流模式
+2025-11-19 11:26:35,020 | INFO     | workflow | info:107 | 🔗 SimpleChatChain 初始化完成
+2025-11-19 11:26:42,252 | INFO     | workflow | info:107 | 🌊 开始流式对话: 2c99f8ef-48a9-4697-999f-c7d6e24d961a
+2025-11-19 11:26:42,253 | ERROR    | workflow | error:115 | ❌ 流式对话失败: LangSmithService.log_model_call() got an unexpected keyword argument 'input_messages'
+2025-11-19 11:27:00,025 | ERROR    | workflow | error:115 | CLI 聊天错误: 
+2025-11-19 14:21:56,343 | INFO     | langsmith_service | info:107 | ✅ LangSmith 监控已启用
+2025-11-19 14:21:56,452 | INFO     | ollama_service | info:107 | ✅ Ollama模型初始化成功
+2025-11-19 14:21:56,452 | INFO     | ollama_service | info:107 | 路由模型: qwen3:8b
+2025-11-19 14:21:56,452 | INFO     | ollama_service | info:107 | 对话模型: deepseek-v3.1:671b-cloud
+2025-11-19 14:21:56,948 | INFO     | workflow | info:107 | 🔗 SimpleChatChain 初始化完成
+2025-11-19 14:21:56,954 | INFO     | workflow | info:107 | 📊 工作流图构建完成
+2025-11-19 14:21:56,955 | INFO     | workflow | info:107 | 🚀 ChatWorkflow 初始化完成
+2025-11-19 14:21:56,955 | INFO     | workflow | info:107 | 🤖 使用 LangGraph 工作流模式
+2025-11-19 14:22:13,892 | INFO     | workflow | info:107 | 🚀 工作流开始: 788a217a-b62a-4d85-9ef8-067dbdc8a4d5
+2025-11-19 14:22:13,893 | INFO     | workflow | info:107 | 🚀 节点开始: intent_analysis
+2025-11-19 14:22:13,893 | INFO     | workflow | info:107 | 🚀 节点开始: intent_analysis
+2025-11-19 14:22:17,743 | INFO     | workflow | info:107 | 🎯 意图分析: 'helper' -> general_chat (置信度: 0.85)
+2025-11-19 14:22:17,743 | INFO     | workflow | info:107 | ✅ 节点完成: intent_analysis
+2025-11-19 14:22:17,743 | INFO     | workflow | info:107 | 💬 路由到自然对话 (意图: IntentType.GENERAL_CHAT, 置信度: 0.85)
+2025-11-19 14:22:17,743 | INFO     | workflow | info:107 | 🚀 节点开始: natural_chat
+2025-11-19 14:22:17,744 | INFO     | workflow | info:107 | 💬 构建对话消息: 2 条消息
+2025-11-19 14:22:17,744 | INFO     | workflow | info:107 | 🚀 节点开始: chat_generation
+2025-11-19 14:22:20,087 | INFO     | workflow | info:107 | ✅ 节点完成: natural_chat
+2025-11-19 14:22:20,087 | INFO     | workflow | info:107 | ✅ 工作流完成: 788a217a-b62a-4d85-9ef8-067dbdc8a4d5
+2025-11-19 14:24:25,745 | INFO     | workflow | info:107 | 🚀 工作流开始: 788a217a-b62a-4d85-9ef8-067dbdc8a4d5
+2025-11-19 14:24:25,746 | INFO     | workflow | info:107 | 🚀 节点开始: intent_analysis
+2025-11-19 14:24:25,747 | INFO     | workflow | info:107 | 🚀 节点开始: intent_analysis
+2025-11-19 14:24:27,889 | INFO     | workflow | info:107 | 🎯 意图分析: '订单 564875' -> order_inquiry (置信度: 0.95)
+2025-11-19 14:24:27,921 | INFO     | workflow | info:107 | ✅ 节点完成: intent_analysis
+2025-11-19 14:24:27,921 | INFO     | workflow | info:107 | 🎯 路由到订单诊断 (置信度: 0.95)
+2025-11-19 14:24:27,921 | INFO     | workflow | info:107 | 🚀 节点开始: order_diagnosis
+2025-11-19 14:24:28,429 | INFO     | workflow | info:107 | 📦 获取订单信息: 564875
+2025-11-19 14:24:28,933 | INFO     | workflow | info:107 | 🚀 节点开始: diagnosis_thinking
+2025-11-19 14:24:42,417 | INFO     | workflow | info:107 | 🔍 生成诊断结果: 564875
+2025-11-19 14:24:42,418 | INFO     | workflow | info:107 | 🚀 节点开始: chat_generation
+2025-11-19 14:24:48,316 | INFO     | workflow | info:107 | 💬 生成最终回复: 564875
+2025-11-19 14:24:48,319 | INFO     | workflow | info:107 | ✅ 节点完成: order_diagnosis
+2025-11-19 14:24:48,319 | INFO     | workflow | info:107 | ✅ 工作流完成: 788a217a-b62a-4d85-9ef8-067dbdc8a4d5
+2025-11-19 14:30:27,321 | ERROR    | workflow | error:115 | CLI 聊天错误: 

课件资料/代码.md

@@ -0,0 +1,174 @@
+# 代码示例
+
+langchain生态
+
+- Chain
+
+```python
+async def main():
+    prompt = ChatPromptTemplate.from_messages([
+        ("system", "你是一个友好、专业的AI助手。"),
+        ("human", "{input}")
+    ])
+
+    # 2) 定义模型（按你本地 Ollama 配置调整）
+    model = ChatOllama(
+        model="qwen3:8b",            # 或你的实际模型名
+        base_url="http://localhost:11434"
+    )
+
+    # 3) 定义输出解析器，将 AIMessage 转字符串
+    parser = StrOutputParser()
+
+    # 4) 组合为表达式管道
+    chain = prompt | model | parser
+
+    # 非流式调用
+    text = await chain.ainvoke({"input": "请用一句话做自我介绍"})
+    print("非流式：", text)
+
+    # 流式调用
+    print("流式：", end="", flush=True)
+    async for chunk in chain.astream({"input": "讲一个简短的笑话"}):
+        print(chunk, end="", flush=True)
+```
+
+- Graph
+
+```python
+# 创建状态图
+async def main():
+    workflow = StateGraph(WorkflowState)
+    
+    # 添加节点
+    workflow.add_node("intent_analysis", self._intent_analysis_wrapper)
+    workflow.add_node("order_diagnosis", self._order_diagnosis_wrapper)
+    workflow.add_node("natural_chat", self._natural_chat_wrapper)
+    
+    # 设置入口点
+    workflow.set_entry_point("intent_analysis")
+    
+    # 添加条件边
+    workflow.add_conditional_edges(
+        "intent_analysis",
+        self._route_after_intent_analysis,
+        {
+            "order_diagnosis": "order_diagnosis",
+            "natural_chat": "natural_chat",
+            "end": END
+        }
+    )
+    
+    # 添加结束边
+    workflow.add_edge("order_diagnosis", END)
+    workflow.add_edge("natural_chat", END)
+    
+    # 编译图
+    agent = workflow.compile()
+
+    messages = [HumanMessage(content="Add 3 and 4.")]
+    messages = agent.invoke({"messages": messages})
+    for m in messages["messages"]:
+        print(m.content)
+```
+
+对应流程图（Mermaid）
+
+```mermaid
+flowchart TD
+    IA[intent_analysis] --> ROUTE{判断意图}
+    ROUTE -->|order_diagnosis| OD[order_diagnosis]
+    ROUTE -->|natural_chat| NC[natural_chat]
+    ROUTE -->|end| END((END))
+    
+    OD --> END
+    NC --> END
+```
+
+说明：当意图分析结果为空或不满足条件时，视为“否”分支，进入 `natural_chat`。
+
+- Agent
+
+```python
+from langchain.agents import create_agent
+
+def get_weather(city: str) -> str:
+    """Get weather for a given city."""
+    return f"It's always sunny in {city}!"
+
+agent = create_agent(
+    model="claude-sonnet-4-5-20250929",
+    tools=[get_weather],
+    system_prompt="You are a helpful assistant",
+)
+
+# Run the agent
+agent.invoke(
+    {"messages": [{"role": "user", "content": "what is the weather in sf"}]}
+)
+```
+
+Go生态
+
+- chain
+  
+```go
+package main
+
+import (
+    "context"
+    "fmt"
+    "os"
+
+    "github.com/cloudwego/eino/compose"
+    "github.com/cloudwego/eino/components/prompt"
+    "github.com/cloudwego/eino/schema"
+    "github.com/cloudwego/eino-ext/components/model/openai"
+)
+
+func main() {
+    ctx := context.Background()
+
+    // 1) 定义 ChatTemplate（输入键：query），使用 Jinja2 格式
+    pt := prompt.FromMessages(
+        schema.Jinja2,
+        schema.SystemMessage("你是一个友好的AI助手。请用中文回答。"),
+        schema.UserMessage("用户问题：{{query}}"),
+    )
+
+    // 2) 创建 ChatModel（以 OpenAI 为例，按需替换为其他提供商）
+    cm, err := openai.NewChatModel(ctx, &openai.ChatModelConfig{
+        APIKey: os.Getenv("OPENAI_API_KEY"),
+        BaseURL: os.Getenv("OPENAI_BASE_URL"), // 可选，自建或代理时设置
+        Model:   os.Getenv("MODEL_NAME"),       // 例如："gpt-4o-mini"
+    })
+    if err != nil {
+        fmt.Println("初始化模型失败:", err)
+        return
+    }
+
+    // 3) 组合并编译 Chain：模板 -> 模型
+    chain, err := compose.NewChain[map[string]any, *schema.Message]().
+        AppendChatTemplate(pt). // prompt
+        AppendChatModel(cm). // model
+        // AppendBranch 分支
+        // AppendPassthrough  空透传节点
+        // AppendParallel 并行节点
+        // AppendGraph 子图节点
+        // AppendLambda 自定义函数节点
+        Compile(ctx) // 编译链
+    if err != nil {
+        fmt.Println("编译链失败:", err)
+        return
+    }
+
+    // 4) 调用链（非流式）
+    msg, err := chain.Invoke(ctx, map[string]any{"query": "简要介绍一下 Eino 的 Chain"})
+    if err != nil {
+        fmt.Println("调用失败:", err)
+        return
+    }
+
+    fmt.Println("回复：", msg.Content)
+}
+```