dive-into-langgraph

A comprehensive guide and reference for building agents using LangGraph 1.0, including ReAct agents, state graphs, and tool integrations.

Security Scan

The skill content and included code legitimately implement a LangGraph guide and examples, but the runtime instructions request API keys/env vars (for model providers) while the registry metadata does not declare any required credentials — an incoherence you should review before installing.

Purpose & Capability

The name/description (LangGraph guide, ReAct agents, state graphs, tools) matches the files and examples included. The Python examples and tools are consistent with teaching/demo usage of LangGraph and LangChain.

Instruction Scope

SKILL.md instructs users to install many LangGraph/LangChain-related Python packages and to populate a .env with model provider credentials (DASHSCOPE_*, ARK_*). The runtime examples load .env and create LLMs using DASHSCOPE_BASE_URL / DASHSCOPE_API_KEY. This is in-scope for a LangGraph tutorial, but the instructions explicitly require secret API keys and network endpoints — so review before supplying secrets.

Install Mechanism

There is no install spec in the registry (instruction-only), and the SKILL.md shows standard pip installation commands. No downloads from untrusted URLs or archive extraction are present. The dependency list is broad but appropriate for a LangGraph tutorial.

Credentials

The skill expects model-provider environment variables (e.g., DASHSCOPE_BASE_URL, DASHSCOPE_API_KEY, ARK_BASE_URL, ARK_API_KEY) and the code reads these via os.getenv. However, the registry metadata lists no required env vars or primary credential — a mismatch. Requesting API keys is reasonable for examples that call external model providers, but the registry should declare this explicitly so users know what secrets will be needed.

Persistence & Privilege

The skill does not request permanent inclusion (always:false), does not write system-wide configs, and only loads .env (dotenv) for credentials. It does not modify other skills or agent settings. Autonomous invocation (model invocation enabled) is the platform default and not an additional risk here.

What to consider before installing

This skill is a tutorial and includes example code that will call external LLM endpoints. Before installing or running it: 1) Expect to provide model API keys (e.g., DASHSCOPE_*, ARK_*) in a .env — the registry metadata did not declare these, so don't paste secrets unless you're comfortable. 2) Review the included examples and tools (they appear benign; math/safe_eval uses a safe AST evaluator). 3) Run examples in an isolated environment/virtualenv and avoid exposing production credentials. 4) If you plan to allow autonomous agents to use this skill, be cautious: providing long-lived model keys to a skill gives it network access to those services. 5) If anything is unclear, ask the skill author for a clear list of required env vars and for the upstream repository/source (SKILL.md links point to a GitHub repo and online docs — verify those sources).

LangGraph 是由 LangChain 团队开发的开源 Agent 框架。v1.0 是稳定版本，框架能力全面升级，支持中间件、状态图、多智能体等高级功能。本 skill 内容由《LangGraph 1.0 完全指南》提供。

LangGraph 1.0 完全指南：

• 在线文档：https://luochang212.github.io/dive-into-langgraph/
• GitHub：https://github.com/luochang212/dive-into-langgraph

安装依赖

基础依赖：

pip install \
  langgraph \
  "langchain[openai]" \
  langchain-community \
  langchain-mcp-adapters \
  python-dotenv \
  pydantic

环境变量

使用模型供应商的大模型需要设置环境变量，推荐使用阿里云百炼（DashScope）的模型：

# 阿里云百炼 (DashScope)
# 获取地址: https://bailian.console.aliyun.com/
DASHSCOPE_BASE_URL=https://dashscope.aliyuncs.com/compatible-mode/v1
DASHSCOPE_API_KEY=your_api_key_here

# 火山方舟 (ARK)
# 获取地址: https://console.volcengine.com/ark/
ARK_BASE_URL=https://ark.cn-beijing.volces.com/api/v3
ARK_API_KEY=your_api_key_here

# 或者其他模型供应商...

请将环境变量添加到 .env 文件，并填入你的 API key。

章节概览

序号	章节	内容简介	在线阅读	离线阅读
1	快速入门	创建你的第一个 ReAct Agent	在线	本地
2	状态图	使用 StateGraph 创建工作流	在线	本地
3	中间件	预算控制、消息截断、敏感词过滤、PII 检测	在线	本地
4	人机交互	使用 HITL 中间件实现人机交互	在线	本地
5	记忆	短期记忆、长期记忆	在线	本地
6	上下文工程	使用 State、Store、Runtime 管理上下文	在线	本地
7	MCP Server	创建 MCP Server 并接入 LangGraph	在线	本地
8	监督者模式	两种方法：tool-calling、langgraph-supervisor	在线	本地
9	并行	节点并发、@task 装饰器、Map-reduce、Sub-graphs	在线	本地
10	RAG	向量检索、关键词检索、混合检索	在线	本地
11	网络搜索	DashScope、Tavily 和 DDGS	在线	本地

官方资源

• LangChain 官方文档
• LangGraph 官方文档
• Deep Agents 官方文档
• LangMem 官方文档
• LangChain GitHub 仓库
• LangGraph GitHub 仓库
• langchain-academy GitHub 仓库