A Git Branch for Every Agent Pattern

What it does This repository is the code companion to a Udemy course on LangGraph. It collects several agent projects—Agentic RAG, ReAct loops, and reflection-based agents—each isolated in its own branch and broken into bite-sized commits. Think of it as a textbook that happens to be a Git repo.

The interesting bit The pedagogy is in the version history. You are meant to step through commits to watch an Agentic RAG pipeline grow from folder scaffolding to self-critique and adaptive routing. Using Git itself as the lesson plan is a neat structural hack.

Key highlights

• Five project branches: Agentic RAG, ReAct agent, reflection demo, reflection agent, and reflexion agent with memory of past runs.
• The Agentic RAG branch includes 11 explicit lesson commits covering document grading, Tavily web search, and conditional graph wiring.
• Examples require live API keys (OpenAI and optionally Tavily and LangSmith) to execute.
• Includes community access and a dedicated job board, LangJobs.dev.

Caveats

• Several branch links in the README contain a typo (langgaph-course), so manual URL correction may be needed.
• The content is coursework, not a reusable library; expect to enroll in the Udemy course for full context.

Verdict Grab this if you are a developer who wants a guided, commit-by-commit dissection of LangGraph patterns. Skip it if you need a standalone framework or drop-in utilities.