LangChain 2026 生态全景：从 LangGraph 到 LangSmith

from langgraph.graph import StateGraph, END
from typing import TypedDict, Annotated
from langgraph.graph.message import add_messages

class AgentState(TypedDict):
    messages: Annotated[list, add_messages]
    context: dict
    iteration: int

def research_node(state: AgentState):
    # 调用搜索工具获取信息
    result = search_tool.invoke(state["messages"][-1].content)
    return {"messages": [{"role": "tool", "content": result}]}

def reasoning_node(state: AgentState):
    # LLM 推理节点
    response = llm.invoke(state["messages"])
    return {"messages": [response]}

def should_continue(state: AgentState) -> str:
    last_msg = state["messages"][-1]
    if state["iteration"] >= 5:
        return END
    if last_msg.tool_calls:
        return "tools"
    return END

# 构建工作流图
workflow = StateGraph(AgentState)
workflow.add_node("research", research_node)
workflow.add_node("reasoning", reasoning_node)
workflow.add_node("tools", tool_executor)

workflow.set_entry_point("reasoning")
workflow.add_conditional_edges("reasoning", should_continue)
workflow.add_edge("tools", "reasoning")

app = workflow.compile(checkpointer=MemorySaver())

from langsmith import Client, traceable

client = Client()

@traceable(name="research_agent", tags=["production"])
async def research_agent(query: str):
    # 自动上报 trace 到 LangSmith
    docs = await retriever.ainvoke(query)
    answer = await llm.ainvoke(messages)
    
    # 记录评估数据
    client.create_feedback(
        run_id=run.id,
        key="answer_quality",
        score=0.85,
        comment="Good coverage of the topic"
    )
    return answer

from langserve import add_routes
from fastapi import FastAPI

app = FastAPI()

# 添加 Agent 路由，自动生成 REST API
add_routes(
    app,
    agent_app,
    path="/api/research-agent",
    input_type=AgentInput,
    output_type=AgentOutput,
    enable_feedback_endpoint=True,
    enable_trace_endpoint=True,
)

# 自动提供 /invoke, /stream, /batch 端点
# 自动生成 /playground 交互界面

LangChain 2026 生态
├── 核心框架
│   ├── langchain-core      # 基础抽象层
│   ├── langchain           # 集成与工具
│   ├── langgraph           # 状态机 Agent 编排
│   └── langgraph-sdk       # 多语言 SDK
├── 开发工具
│   ├── LangSmith           # 可观测性 + 评估
│   ├── LangServe           # 服务部署
│   ├── LangChain CLI       # 项目脚手架
│   └── LangChain Hub       # Prompt/组件市场
├── 集成生态（200+ 集成）
│   ├── LLM 提供商（OpenAI, Anthropic, Google, Meta...）
│   ├── 向量数据库（Pinecone, Weaviate, Qdrant, Chroma...）
│   ├── 文档加载器（50+ 格式）
│   └── 工具生态（搜索、代码执行、API 调用...）
└── 行业方案
    ├── langchain-banking   # 金融行业模板
    ├── langchain-healthcare # 医疗合规模板
    └── langchain-legal     # 法律分析模板