LangGraph CLI 快速入门

什么是 LangGraph CLI

LangGraph CLI 是一个多平台命令行工具，用于在本地构建和运行 LangGraph API 服务器。生成的服务器包含图的所有运行、线程、助手等的 API 端点，以及运行代理所需的其他服务，包括用于检查点和存储的托管数据库。

---

环境准备

Python 版本要求

Python >= 3.11

创建虚拟环境

# 安装虚拟环境库
pip install virtualenv

# 创建虚拟环境
virtualenv env_name

# 激活虚拟环境（Windows）
env_name\Scripts\activate

# 激活虚拟环境（Linux/Mac）
source env_name/bin/activate

# 退出虚拟环境
deactivate

---

安装与创建应用

安装 LangGraph CLI

pip install --upgrade "langgraph-cli[inmem]"

创建新应用

langgraph new path/to/your/app --template new-langgraph-project-python

如果未指定模板，会显示交互式菜单供选择。

(fastapi_env) C:\Users\21129\Documents\百度网盘\langgraphAgent>langgraph new langgraph-demo
🌟 Please select a template:
1. New LangGraph Project - A simple, minimal chatbot with memory.
2. ReAct Agent - A simple agent that can be flexibly extended to many tools.
3. Memory Agent - A ReAct-style agent with an additional tool to store memories for use across conversational threads.
4. Retrieval Agent - An agent that includes a retrieval-based question-answering system.
5. Data-enrichment Agent - An agent that performs web searches and organizes its findings into a structured format.
Enter the number of your template choice (default is 1): 1

You selected: New LangGraph Project - A simple, minimal chatbot with memory.
Choose language (1 for Python 🐍, 2 for JS/TS 🌐): 1
📥 Attempting to download repository as a ZIP archive...
URL: https://github.com/langchain-ai/new-langgraph-project/archive/refs/heads/main.zip
✅ Downloaded and extracted repository to C:\Users\21129\Documents\百度网盘\langgraphAgent\langgraph-demo
🎉 New project created at C:\Users\21129\Documents\百度网盘\langgraphAgent\langgraph-demo

langgraph.json 是整个项目的配置文件。

{
  "$schema": "https://langgra.ph/schema.json",
  "dependencies": ["."],  // 当前目录
  "graphs": { // 智能体对象
    "agent": "./src/agent/graph.py:graph"
  },
  "env": ".env", // 环境变量 env
  "image_distro": "wolfi" // docker 镜像导出方式
}

安装项目依赖

cd path/to/your/app
pip install -e .

注意：LangGraph 使用 pyproject.toml 替代传统的 setup.py 和 requirements.txt

---

配置 Agent

编辑 graph.py 文件，配置您的大模型和工具：

import os

from langchain_classic.agents import create_react_agent
from langchain_openai import ChatOpenAI
from dotenv import load_dotenv
load_dotenv()
OPENAI_API_KEY = os.getenv('OPENAI_API_KEY')
llm = ChatOpenAI(
    temperature=0.5,
    model='gpt-4o-mini',
    api_key=OPENAI_API_KEY,
    base_url='https://xiaoai.plus/v1'
)

def get_weather(city: str) -> str:
    """Get weather for a given city."""
    return f"城市： {city}，今天的天气晴朗，气温在 28 摄氏度!"

graph = create_react_agent(
    llm,
    tools=[get_weather],
    prompt="你是智能助手！"
)

---

启动服务器

langgraph dev

langgraph dev 以内存模式启动服务器，适用于开发和测试。生产环境需要配置持久存储后端。

启动后可通过 LangGraph Studio（本地 UI）访问和调试您的 Agent。

langgraph dev 选项

选项	默认值	说明
-c, --config	langgraph.json	配置文件路径（声明依赖项、图和环境变量）
--host	127.0.0.1	服务器绑定的主机
--port	2024	服务器绑定的端口
--no-reload	-	禁用自动重载
--n-jobs-per-worker	10	每个工作进程的作业数
--debug-port	-	调试器监听端口
--wait-for-client	FALSE	等待调试器客户端连接后再启动服务器
--no-browser	-	启动时跳过自动打开浏览器
--studio-url	https://smith.langchain.com	LangGraph Studio 实例 URL
--allow-blocking	FALSE	允许同步 I/O 阻塞操作
--tunnel	-	通过 Cloudflare 公共隧道暴露本地服务器

╦  ┌─┐┌┐┌┌─┐╔═╗┬─┐┌─┐┌─┐┬ ┬
║  ├─┤││││ ┬║ ╦├┬┘├─┤├─┘├─┤
╩═╝┴ ┴┘└┘└─┘╚═╝┴└─┴ ┴┴  ┴ ┴

- 🚀 API: http://127.0.0.1:2024
- 🎨 Studio UI: https://smith.langchain.com/studio/?baseUrl=http://127.0.0.1:2024
- 📚 API Docs: http://127.0.0.1:2024/docs

This in-memory server is designed for development and testing.
For production use, please use LangSmith Deployment.

常用示例：

# 修改端口
langgraph dev --port 3000

# 禁用自动重载（生产模式）
langgraph dev --no-reload

# 通过隧道暴露（解决 localhost 连接问题）
langgraph dev --tunnel

# 禁用浏览器自动打开
langgraph dev --no-browser

---

测试方式

Python SDK

pip install langgraph-sdk

异步方式：

from langgraph_sdk import get_client
import asyncio

client = get_client(url="http://localhost:2024")

async def main():
    async for chunk in client.runs.stream(
        None,  # Threadless run
        "agent",  # Name of assistant
        input={
            "messages": [{
                "role": "human",
                "content": "What is LangGraph?",
            }],
        },
    ):
        print(f"Event: {chunk.event}")
        print(chunk.data)

asyncio.run(main())

同步方式：

from langgraph_sdk import get_sync_client

client = get_sync_client(url="http://localhost:2024")

for chunk in client.runs.stream(
    None,
    "agent",
    input={"messages": [{"role": "human", "content": "What is LangGraph?"}]},
    stream_mode="messages-tuple",
):
    print(f"Event: {chunk.event}")
    print(chunk.data)

JavaScript SDK

npm install @langchain/langgraph-sdk

const { Client } = await import("@langchain/langgraph-sdk");

const client = new Client({ apiUrl: "http://localhost:2024"});

const streamResponse = client.runs.stream(
    null,
    "agent",
    {
        input: { "messages": [{ "role": "user", "content": "What is LangGraph?"}] },
        streamMode: "messages-tuple",
    }
);

for await (const chunk of streamResponse) {
    console.log(`Event: ${chunk.event}`);
    console.log(JSON.stringify(chunk.data));
}

REST API

curl -s --request POST \
    --url "http://localhost:2024/runs/stream" \
    --header 'Content-Type: application/json' \
    --data '{
        "assistant_id": "agent",
        "input": {
            "messages": [{"role": "human", "content": "What is LangGraph?"}]
        },
        "stream_mode": "messages-tuple"
    }'

---

速记要点

步骤	命令	关键点
创建环境	virtualenv env_name	Python >= 3.11
安装 CLI	pip install langgraph-cli[inmem]	带 inmem 后缀
创建项目	langgraph new --template python	选模板
安装依赖	pip install -e .	编辑模式
启动服务	langgraph dev	内存模式
测试 API	SDK / REST	四种方式

核心一句话

LangGraph CLI = 一套命令 + 一个图 + 多个 API

记忆口诀

• 安装：pip install langgraph-cli
• 创建：langgraph new
• 启动：langgraph dev
• 测试：四种方式（Studio / Python / JS / REST）