10 interactive labs. From "why do I need this?" to building real agent workflows.
Manual vs LangGraph — feel the pain, see the solution
What goes wrong when you build agents with raw LLM API calls
Drag, drop, and build graphs visually
START, LLM, Tool, END — the building blocks
Fixed edges vs conditional edges
The shared memory that flows through the graph
Assemble a 4-node graph step by step
Watch nodes get revisited — the agent superpower
Every loop needs a leash
ReAct, Reflection, Planning — all as graphs