Agent-to-Agent Protocols (A2A)

About this workflow

Connect two agents over an explicit protocol - task handoff, acknowledgment, and failure handling - instead of hoping chat works out.

What happens when you run it

1. 1
   Write the message schemas and state machine before any agent code.

   Runs flow-agent-to-agent-protocols-a2a-1-define-the-protocol on Claude Sonnet 4.6
2. 2
   Reliable-enough transport with full observability.

   Runs flow-agent-to-agent-protocols-a2a-2-build-the-message-layer on Claude Sonnet 4.6
3. 3
   Each agent's loop stays private; the protocol is its public face.

   Runs flow-agent-to-agent-protocols-a2a-3-wrap-agents-as-protocol-endpoints on Claude Sonnet 4.6
4. 4
   The demo that matters: cooperation surviving imperfection.

   Runs flow-agent-to-agent-protocols-a2a-4-run-the-revision-loop-with-failure-injection on Claude Sonnet 4.6

Best for

• •Recurring tasks you currently copy-paste between tools.
• •When a single prompt isn't enough — you need a chain where output(n) feeds prompt(n+1).
• •Fork and adapt if your inputs differ materially from the author's use case.

Models used