LangGraph Starter Kit
A scaffolding CLI and starter kit for building multi-agent applications with LangGraph, offering seven agent patterns, six LLM providers and an HTTP server with streaming out of the box.
LangGraph Starter Kit gives a project the boilerplate that multi-agent LangGraph applications tend to repeat: supervisors, handoff tools, provider configuration and persistence. A command-line scaffolder generates a ready-to-run project with a chosen LLM provider and set of agent patterns, including swarm, supervisor, human-in-the-loop, structured output, research, retrieval-augmented generation and customer support.
It is built on LangChain's createAgent API with its subagents and handoffs patterns, supports OpenAI, Anthropic, Google, Groq, DeepSeek and local Ollama as providers, and includes an HTTP server built on Fastify with invoke, streaming, resume and thread-history endpoints. Generated projects come with tests, TypeScript types, LangGraph Studio configuration for visual debugging, optional LangSmith tracing, in-memory or PostgreSQL persistence, and Docker Compose and CI configuration.