Python LangGraph Guides

These guides walk through an incremental Python integration path for LangGraph developers:

Start from a minimal stateless LangGraph chatbot
Turn on memory-service features one by one
Keep code changes small and explicit using checkpoint-based diffs

Each guide references runnable checkpoints under python/examples/langgraph/doc-checkpoints/ and includes site-tests-backed curl scenarios.

Prerequisites

Before starting, complete LangGraph Dev Setup and ensure Memory Service + Keycloak are running. Also complete Step 2 on that page (build local memory-service-langchain wheel + UV_FIND_LINKS); this is temporary until the package is released.

Tutorial Path

Getting Started

Build a minimal LangGraph chatbot, then enable memory-backed conversations with persistent checkpointing.

Conversation History

Record USER/AI turns in the history channel and expose read APIs.

Indexing and Search

Add indexed history content and conversation search.

Conversation Forking

Pass fork metadata and list conversation forks.

Response Recording and Resumption

Stream responses and support resume-check, resume, and cancel.

Add memberships and ownership transfer APIs.

Episodic Memories

Use the LangGraph BaseStore backend to give your agent persistent per-user memories.

Reference

Client Configuration

Environment variables, explicit config, and from_env() factory usage.

Dev Setup

Reproducible uv workflow, Dockerized dependencies, and fast Python-only docs tests.