LangGraph¶

A library for building stateful, multi-actor applications with LLMs using directed graphs.

Installation¶

pip install langgraph langchain-openai

Core concepts¶

Concept	Description
State	A typed dict shared across all nodes
Node	A function that reads and writes to state
Edge	A connection between nodes (can be conditional)
Graph	The assembled pipeline

Minimal ReAct agent¶

from langgraph.prebuilt import create_react_agent
from langchain_openai import ChatOpenAI
from langchain_core.tools import tool

@tool
def get_weather(city: str) -> str:
    """Get current weather for a city."""
    return f"Sunny, 22°C in {city}"

model = ChatOpenAI(model="gpt-4o-mini")
agent = create_react_agent(model, tools=[get_weather])

result = agent.invoke({"messages": [("user", "What's the weather in Paris?")]})
print(result["messages"][-1].content)

Building a custom graph¶

from typing import TypedDict, Annotated
from langgraph.graph import StateGraph, END
import operator

class AgentState(TypedDict):
    messages: Annotated[list, operator.add]

def call_model(state: AgentState):
    # ... call LLM ...
    return {"messages": [response]}

def should_continue(state: AgentState):
    last = state["messages"][-1]
    return "tools" if last.tool_calls else END

graph = StateGraph(AgentState)
graph.add_node("agent", call_model)
graph.set_entry_point("agent")
graph.add_conditional_edges("agent", should_continue)
app = graph.compile()

Persistence & human-in-the-loop¶

LangGraph supports checkpointing so agents can pause and resume:

from langgraph.checkpoint.memory import MemorySaver

checkpointer = MemorySaver()
app = graph.compile(checkpointer=checkpointer, interrupt_before=["tools"])

Notes & experiments¶

Add your LangGraph notes and experiment results here.