Skip to main content
skill

langgraph · davila7 — Claude Code skill: build agents as graphs

A Claude skill for building LangGraph agents: state, branching, checkpointers and the ReAct loop.

Install into Claude Code · claude-code-templates All methods and platforms ↓
npx claude-code-templates@latest --skill ai-research/langgraph
Open on GitHub

What langgraph does

This skill turns Claude into a LangGraph architect. It walks through StateGraph construction: declaring state, adding nodes and edges, wiring conditional routing. Reducers get their own treatment — add_messages appends instead of overwriting, and a custom reducer can merge dicts coming from several nodes.

Inside are code templates: a basic ReAct agent with ToolNode, state with multiple reducers, branching on query type. There is also an anti-patterns section. A loop with no exit burns tokens, and stateless nodes throw away the main LangGraph benefit — you cannot resume a conversation.

You need Python 3.9+, the langgraph package and LLM API access such as OpenAI or Anthropic. If your graph has a cycle, put an iteration counter in state: without an exit condition the agent loops and hits your request limit.

What it helps with

  • build a ReAct agent that calls tools
  • define state with reducers for accumulating data
  • route queries into branches with conditional edges
  • add a checkpointer and resume a conversation
  • catch an infinite loop in an agent graph

Example prompts for the agent

  • help me build a StateGraph for an agent with search and a calculator
  • write state with a reducer that merges findings from several nodes
  • add an exit condition based on an iteration counter

How to install langgraph

1claude-code-templates

npx claude-code-templates@latest --skill ai-research/langgraph

2Skills CLI (Vercel)

npx skills add davila7/claude-code-templates --skill langgraph -a claude-code

3Manually: put the files in place

~/.claude/skills/langgraph/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/langgraph/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about langgraph

How do I install langgraph in Claude?

Use the templates: npx claude-code-templates@latest --skill ai-research/langgraph. Or run npx skills add davila7/claude-code-templates --skill langgraph -a claude-code.

What does the langgraph skill do?

It helps you design LangGraph agents: graphs, state, reducers, branching, checkpointers and human-in-the-loop patterns. Code templates and common mistakes are included.

What do I need to run it?

Python 3.9+, the langgraph package installed, and LLM API access — Anthropic, Anthropic or another provider. Some familiarity with graph concepts helps.

Similar skills

Everything for AI and ML →

langgraph by other authors comes first.

langgraph sickn33 A skill for building production-grade agents with deliberate structure. Design graph topology, manage state, add cycles and branches, persist progress via checkpoints, and loop humans back into decisions. CX Codex +1 1,300installs langgraph OrchestKit A skill with 41 rules for LangGraph 1.x in Python: state schemas, node timeouts, routing, supervisors, checkpoints. Useful when you build a pipeline or a multi-agent system. CX Codex +1 359installs langgraph-persistence langchain-ai A Claude Code skill for LangGraph persistence: wiring a checkpointer, keeping conversations apart by thread_id, and replaying older graph state. CX Codex +1 18,000installs langgraph-python-quickstart langchain-ai Scaffolds a minimal local LangGraph Python agent using the official quickstart. Useful when you need to test a Graph API agent in an isolated folder. CX Codex +1 5,200installs options-analytics-agent-langgraph reason-machines Builds a LangGraph agent that pulls options data from Polygon.io, caches it in ChromaDB and keeps memory across sessions. Useful when you want your own trading assistant. CX Codex +1 2,000installs langgraph-code-review existential-birds LangGraph code review skill for Claude Code and coding agents. It flags state mutations, missing reducers, and broken graph topologies. CX Codex +1 884installs langchain-development Mindrally Gives your agent a rulebook for LangChain and LangGraph work in Python: chain composition, agents, memory, RAG. Useful once a Python LLM project outgrows one file. CX Codex +1 681installs migrating-langchain-to-pydantic-ai pydantic A skill that moves Python apps off LangChain, LangGraph and Deep Agents onto Pydantic AI. Useful when you want to keep the behavior and drop the framework shape. CX Codex +1 314installs

From the same set224

The whole set →

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor