The newcomer — Start here if multi-agent systems are new to you.

Steps

1. 1

   Why multi-agent systems exist

   Ground the rest in the problem MAS is trying to solve before meeting any solution.

   You should be able to explain in one sentence why a single LLM call isn't enough for the tasks this site is about.

2. 2

   What an agent is made of

   Pin down what composes a single agent before the patterns rely on it — the parts and the Reason → Act → Observe loop that drives them.

   You should be able to name the parts that compose a single agent and describe the Reason → Act → Observe loop in one sentence.

3. 3

   The three layers of the stack

   Place each later concept on the correct layer so the field stops feeling like a flat soup of buzzwords.

   You should be able to place 'tool calling', 'reducer', and 'orchestration' on the correct layer.

4. 4

   The ladder of agency

   See the gradient from a workflow to a fully autonomous agent so the trade-off is concrete, not philosophical.

   You should be able to point at where on the ladder a system needs to sit before the word 'agent' is justified.

5. 5

   The six coordination patterns

   Most production work re-instantiates one of six shapes; recognising them early prevents reinventing them.

   You should be able to list at least four of the six and describe each in one sentence.

6. 6

   Run the simplest pattern

   Concrete, runnable, and small enough that nothing in the page surprises you. Builds confidence the rest is mechanical.

   You should be able to describe what the orchestrator node does and what it does not do.