LangGraph Basics
Control the execution flow of your agents
The libraries we've used so far, like CrewAI, are great for building agents that use linear (or hierarchical) flow of execution. However, those are the only available options. For example, what if you want to decide the order of execution or create a loop in your workflow? This is where LangGraph1 becomes essential.
Tutorial Goals
In this tutorial, you will:
- Learn what LangGraph is
- Build tools that your LangGraph app can use
- Create a graph that controls the flow of your application
- Create state that will be passed between the graph nodes
- Chat with the graph to see how it works
We'll reuse the habit tracker from the LLM function calling tutorial to build a LangGraph app. This app will allow you to add, complete, and view habits. Additionally, we'll create tools that the LangGraph app can use to interact with the habit tracker.