Dive into LangGraph skill
A LangGraph 1.0 guide as a skill: ReAct agents, StateGraph, middleware, memory, MCP and RAG, so Claude Code writes LangChain code without outdated APIs
- 461 ★ on the repository
- Last commit
Minimal access
Why this level
- The skill is reference chapters and examples and runs nothing by itself
- Example code calls paid model provider APIs with your key when you run it yourself
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/dive-into-langgraphDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill is reference chapters and examples and runs nothing by itself
- Example code calls paid model provider APIs with your key when you run it yourself
This is third-party code. Review the repository files before installing.
What it does
Dive into LangGraph is an open guide, "LangGraph 1.0 完全指南", in Chinese that since March 2026 also ships as an agent skill. The skill gives Claude Code current practice for LangChain and LangGraph 1.0, so the agent does not lean on 0.6 interfaces, which the material leaves out. SKILL.md leads through eleven chapters: a first ReAct agent, StateGraph workflows, middleware for budget, message trimming, word filtering and PII, human in the loop, short- and long-term memory, context handling through State, Store and Runtime, building an MCP server, the supervisor pattern with subagents as tools, parallelism, three ways to do RAG and web search through DashScope, Tavily, DDGS and Exa. Each chapter sits in the references folder for offline reading with a link to the online version. Code examples are collected in scripts/examples.py. The repository also holds full notebooks for 14 chapters, a Gradio app and a small MCP server project. Running the examples needs the dependencies from pyproject.toml (the uv sync command) and a model provider key. SKILL.md recommends Alibaba DashScope, lists Volcano Engine Ark too, and any other provider works.
Who it is for. For Python developers building agents on LangChain and LangGraph 1.0 with Claude Code who can read Chinese technical documentation.
Good fit when
- You need a first ReAct agent on LangGraph 1.0 whose code relies on the current API
- You need middleware to cap budget, trim messages or detect personal data in a conversation
- The agent needs short- and long-term memory or a human in the approval loop
- You must implement a supervisor pattern where subagents are called as tools
- You want to compare vector, keyword and hybrid search for RAG on ready examples
Not a fit when
- You do not read Chinese: SKILL.md and the reference chapters are written in Chinese, and only the README has an English translation
- Your project runs on LangGraph 0.x: the material carries no 0.6 legacy
- You only need a quick choice between LangChain, LangGraph and Deep Agents: the catalog has official LangChain skills for that
- You need a commercial adaptation of the guide: the license forbids commercial use
Example request
Using the dive-into-langgraph skill, write a LangGraph 1.0 agent with memory, a human in the loop and middleware that trims long conversationsLimitations
- Content is licensed CC BY-NC-SA 4.0: non-commercial use only and the same license for derivatives.
- The guide and the skill are in Chinese. The README has an English version in docs/README-en.md.
- Running the examples needs a model provider key in .env, for example DASHSCOPE_API_KEY or ARK_API_KEY, and dependencies through uv sync.
- The MCP section uses the langchain.mcp namespace, which is still marked beta.
- The skill covers 11 chapters. Chapters on Deep Agents, the Gradio app and the langgraph-cli debug page are in the notebooks and the online version.
How to disable. Delete the dive-into-langgraph folder from the agent's skills directory, for example ~/.claude/skills/. Provider variables in your .env stay, so remove them by hand if you no longer need them.
FAQ
How do I install Dive into LangGraph in Claude Code?
Run npx skills add luochang212/dive-into-langgraph. The skill appears in your skills list, and the reference chapters read from the references folder without a network.
What is in the skill and what only in the repository?
The skill holds SKILL.md, 11 reference chapters, example scripts and tutorial tools. The notebooks for 14 chapters, the Gradio app and the MCP server sit alongside in the repository and in the online book.
Which LangGraph versions does the material cover?
Only 1.0: the author states the guide is written entirely for this release and carries no 0.6 legacy.
Do I need an API key for the examples?
Yes. The examples call models through providers that read the key from .env. SKILL.md suggests Alibaba DashScope, lists Volcano Engine Ark too, and any other works.
Can I use the material in a commercial project?
CC BY-NC-SA 4.0 allows non-commercial use only. Contact the author for commercial purposes.
Discussion
No questions or reviews yet.
Related
SkillsOfficial LangChain skills: choose between LangChain, LangGraph and Deep Agents, quickstarts, RAG, memory, human approval and eval engineering
A CLI and MCP server that turns doc sites, repos, PDFs and videos into skills and RAG-ready data
Numman Ali's small plugin marketplace: agent orchestration, open-source maintenance, Gas Town and Z.AI CLI
A universal loader for Anthropic-format skills (SKILL.md) in any coding agent: install, sync and load on demand
LangSmith skills
LangSmith Skills
Official LangChain skills for LangSmith: query and export traces, build eval datasets and write custom evaluators from your agent
A Claude Code reference plus a collection of over 70 skills for development, security, data and docs, with agent configs
README in short
The README opens by announcing that the guide became a skill installed with npx skills add. It then describes the project and version 1.0, dependency setup through uv sync, pre-commit hook setup, a table of 14 chapters linking to notebooks, the langgraph dev debug page, a practical Gradio app, a list of official docs and contribution links. License CC BY-NC-SA 4.0.
README badge
Original SKILL.md
--- name: dive-into-langgraph description: A comprehensive guide and reference for building agents using LangGraph 1.0, including ReAct agents, state graphs, and tool integrations. --- # Dive Into LangGraph 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 在本仓库中同步全部依赖： ```bash uv sync ``` 仓库使用 `pyproject.toml` 声明依赖、`uv.lock` 固定精确版本。MCP 使用 `langchain[mcp]` 内置的 `langchain.mcp.MCPAdapter`。`langchain.mcp` 命名空间当前仍为 **Beta**，本 Skill 只面向当前稳定包中的最新接口，不保留旧形态兼容层。 ## 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： ```bash # 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