Graph I: Hello World Graph

In this section, you will learn the basics of building a LangGraph by constructing a simple “Hello World” graph.

Objectives

Understand and define the AgentState structure.
Create simple node functions to process and update state.
Set up a basic LangGraph structure.
Compile and invoke a LangGraph graph.
Understand how data flows through a single node in LangGraph.

Project Setup

Before writing any code, let’s set up a clean, isolated project environment using uv and install the required dependencies:

1. Initialize the Project

Create a new directory for your project and initialize it with uv:

# Create and enter the project directory
mkdir hello-langgraph
cd hello-langgraph

# Initialize a new project with uv
uv init

2. Install Dependencies

Install the required langgraph package using uv add:

uv add langgraph

This will automatically create a virtual environment (.venv) and lock exact versions in uv.lock. Now you can create a file (e.g., main.py) and start editing!

Graph 0: Hello World

Goal

Build a simple graph that updates a state containing only one variable msg to "Hello, World!".

Sample Input

{"msg": ""}

Sample Output

{"msg": "Hello, World!"}

Plan

Define the state schema (AgentState0) with a single variable msg.
Define the node function hello_node that sets msg to "Hello, World!".
Build and compile the graph.
Invoke the graph with an initial empty message.

Code Implementation

1. Define the State Schema

We define a state schema with a single field msg of type str:

from typing import TypedDict

class AgentState0(TypedDict):
    msg: str

2. Define the Node Function

The node function updates the msg field in the state to "Hello, World!":

def hello_node(state: AgentState0) -> AgentState0:
    state['msg'] = "Hello, World!"
    return state

3. Build and Compile the Graph

We build a single-node graph:

from langgraph.graph import StateGraph, START, END

# Initialize graph
graph0 = StateGraph(AgentState0)

# Add node and transition
graph0.add_node("hello_world", hello_node)
graph0.set_entry_point("hello_world")
graph0.set_finish_point("hello_world")

# Compile
app0 = graph0.compile()

4. Invoke the Graph

We invoke the graph by passing an initial state with an empty string:

result0 = app0.invoke({"msg": ""})
print(result0)
# Output: {'msg': 'Hello, World!'}

Goal

Build a simple graph that takes a user’s name as input and outputs a personalized greeting.

Sample Input

{"name": "Amit"}

Sample Output

{"name": "Amit", "greet": "Hey Amit, how is your day going?"}

Plan

Define the AgentState schema containing name and greet variables.
Define the node function greeting_node that reads the input name and constructs the personalized greeting.
Build and compile the graph.
Invoke the graph with a custom name.

Code Implementation

1. Define the State Schema

The state defines the data structure shared across the graph. We use a TypedDict for this:

from typing import TypedDict

class AgentState(TypedDict):
    name: str
    greet: str

2. Define the Node Function

A node is just a Python function that takes the current state as input and returns an updated dictionary of fields to merge back into the state:

def greeting_node(state: AgentState) -> AgentState:
    state['greet'] = f"Hey {state['name']}, how is your day going?"
    return state

3. Build and Compile the Graph

We use StateGraph to define our nodes and layout:

from langgraph.graph import StateGraph, START, END

# Initialize the graph with the state schema
graph = StateGraph(AgentState)

# Add the node to the graph
graph.add_node("greeter", greeting_node)

# Set the entry point and the finish point
graph.set_entry_point("greeter")
graph.set_finish_point("greeter")

# Compile the graph into a runnable application
app = graph.compile()

4. Invoke the Graph

We invoke the compiled application by passing the initial state:

result = app.invoke({"name": "Amit"})
print(result)
# Output: {'name': 'Amit', 'greet': 'Hey Amit, how is your day going?'}

Exercise: Personalized Compliment Agent 🏗

Goal

Create a Personalized Compliment Agent using LangGraph.

Sample Input

{"name": "Amit"}

Sample Output

"Amit, you're doing an amazing job learning LangGraph!"

Plan

Define a state with name and compliment keys.
Use a single node that updates compliment using the input name.
Build, compile, and invoke the graph.

Solution

from typing import TypedDict
from langgraph.graph import StateGraph, START, END

# 1. State Schema
class AgentState(TypedDict):
    name: str
    compliment: str

# 2. Node Function
def compliment_node(state: AgentState) -> AgentState:
    state['compliment'] = f"{state['name']}, you're doing an amazing job learning LangGraph!"
    return state

# 3. Create Graph
graph = StateGraph(AgentState)
graph.add_node("complimenter", compliment_node)
graph.set_entry_point("complimenter")
graph.set_finish_point("complimenter")

# 4. Compile and Run
app = graph.compile()
result = app.invoke({"name": "Amit"})
print(result['compliment'])

Practice & Exercises

To reinforce what you’ve learned in this section (defining schemas, creating nodes, compiling and invoking graphs), practice with the interactive notebook:

Practice & Exercises

Practice variable state schemas, node functions, building/compiling graphs, and implementing a personalized compliment agent.💻 VS Code | 🚀 Colab | 📥 Download

LangGraph Basics

LangGraph Agents

Graph I: Hello World Graph

Objectives

Project Setup

1. Initialize the Project

2. Install Dependencies

Graph 0: Hello World

Goal

Sample Input

Sample Output

Plan

Code Implementation

1. Define the State Schema

2. Define the Node Function

3. Build and Compile the Graph

4. Invoke the Graph