Explanation
Understanding-oriented. Read these when you want to know why the template is shaped the way it is.
| Page | Question it answers |
|---|---|
| Architecture | How do the A2A handler, LangGraph runtime, and LiteLLM gateway fit together? |
| A2A protocol | What does A2A actually require, and where do naive implementations go wrong? |
| Cost & trace propagation | Why do we emit cost-v1 and parse a2a.trace, and why that specific shape? |
| Output protocol | Why <scratch_pad> / <output> instead of whatever the model emits? |
| LiteLLM gateway | Why route every call through a gateway instead of the provider SDK? |