Module 12 A2A and A2UI Agent Interoperability

🔌 Module 12 · A2A and A2UI: Agent Interoperability

Agents talking to agents, and agents talking to users.

Module 12 of 16 · track Ship · Estimated time: 4–5 hours · Prerequisites: Module 06 and Module 11.

Two protocols solve two different problems. A2A lets agents from different vendors discover and call each other. A2UI gives an agent a way to render real interface, not just text. Both are early, both matter.

Open the module on GitHub ↗: the slides, notebooks, exercises and solutions are there. This page is the reading order, with the tutorial and the workbench alongside.

Read first, on the site

The lesson that frames this module is P1 Design & Spec: Write a Spec an AI Agent Can Build From (8 min): an interface between agents is a spec, and it has the same gate.

What you will be able to do

• Read and publish an A2A agent card
• Make two independently-built agents complete a task together
• Decide when interop beats a direct tool call
• Render an agent-driven interface with A2UI

Concepts in this module

Portable GenAI concepts	AWS-specific surface
Agent-to-agent protocols	A2A on AgentCore
Capability discovery	Strands A2A server
Agent-driven UI	AgentCore Gateway
Interop trade-offs

Portable concepts transfer to any stack. The AWS column is where this module touches the cloud — see docs/concepts/portability-matrix.md.

Run it in this order

#		Step	What it is
1	📖	slides/A2A_Protocol_Strands_AgentCore.pptx	The A2A protocol
2	✏️	exercises/exercise_1_read_the_card.md	Exercise 1 — read the card
3	💻	notebooks/01_A2A_with_Strands.ipynb	A2A with Strands
4	✏️	exercises/exercise_2_two_agent_handshake.md	Exercise 2 — two-agent handshake
5	💻	notebooks/02_A2A_on_AgentCore.ipynb	A2A on AgentCore
6	✏️	exercises/exercise_3_right_tool_right_job.md	Exercise 3 — right tool, right job
7	📖	slides/A2UI_The_UI_Layer_for_AI_Agents.pptx	A2UI — the UI layer
8	💻	notebooks/Notebook_1_A2UI_through_Strands.ipynb	A2UI through Strands
9	💻	notebooks/Notebook_2_A2UI_with_AgentCore.ipynb	A2UI with AgentCore
10	✏️	exercises/A2A_Exercise_Aurora_Grid.md	Aurora Grid — the full A2A exercise
11	✏️	exercises/mini_project_agent_mesh.md	Mini project — agent mesh

📖 read   💻 run   ✏️ practise   📊 workbook   🔖 reference

Worked solutions live in solutions/. Attempt every exercise closed-book first — the solutions are written to be read after you have a wrong answer to compare against.

Common mistakes

• Reaching for A2A when a function call would do. Protocols add latency and failure modes.
• Agent cards that overstate capability. Other agents will believe them.

When it breaks

Every error anyone has hit in this course is in the Error Index, by the exact string. Model access and region problems are in Model and Region Notes. If it is a question rather than an error, check the answered ones first, then ask in Q&A.

Field guide for this module

Frameworks

• Handoff Multiplier — Each protocol hop is a full agent run
• Tool Surface Audit — Trust an agent card like any other tool

Quick reference

• MCP and A2A — When neither is the answer

---

← Module 11 · Amazon Bedrock AgentCore · All modules · Module 13 · Agentic QA and Evaluation →

_{✏️ This page is generated from modules/12-a2a-and-a2ui-interop/README.md. To change it, edit the module README: an edit made here is replaced at the next sync.}