Extending a MCP/A2A Currency Agent with A2UI

Building an Agent with A2A, MCP, ADK, and A2UI This tutorial aims to extend and test a...

title: Extending a MCP/A2A Currency Agent with A2UI published: true series: A2A date: 2026-06-03 21:02:16 UTC tags: aiagent,googleadk,a2aprotocol,a2ui canonical_url: https://xbill999.medium.com/extending-a-mcp-a2a-currency-agent-with-a2ui-dc43ffdc2508

Building an Agent with A2A, MCP, ADK, and A2UI

This tutorial aims to extend and test a currency Agent using A2A and MCP protocols with the A2UI protocol for custom presentation.

Reduce, Re-Use, Re-Cycle!

This paper is a re-visiting of the original currency Agent Codelab:

Getting Started with MCP, ADK and A2A | Google Codelabs

and a GitHub Repo:

GitHub - jackwotherspoon/currency-agent: A sample agent demonstrating A2A + ADK + MCP working together.

In this updated version, the Antigravity CLI is used to add support for A2UI and extend the existing user interface.

What is the A2A protocol?

The Agent2Agent (A2A) protocol, an open communication standard for AI agents, was initially introduced by Google in April 2025. It is specifically engineered to facilitate seamless interoperability within multi-agent systems, enabling AI agents developed by diverse providers or built upon disparate AI agent frameworks to communicate and collaborate effectively.

A good overview of the A2A protocol can be found here:

A2A Protocol

Language Support For the A2A Protocol

The official ADK for Python, GO, and Java provide built-in support for working with the A2A protocol. For other programming languages like JS, Rust, and .NET — 3rd party libraries are available to add support for the protocol.

The main source for A2A Language support is the GitHub repo:

GitHub - a2aproject/A2A: An open protocol enabling communication and interoperability between opaque agentic applications.

A2UI

A2UI (Agent-to-User Interface) is an open-source protocol that allows AI agents to dynamically generate and stream rich, interactive user interfaces in real-time. [1, 2]

Instead of an AI relying on pre-built screens or just returning plain text in a chat window, A2UI enables the agent to instantly build and display tailored components like interactive charts, date-pickers, or approval forms. [1, 2]

What is A2UI? - A2UI

More Word Salad Protocols — What about A2A-XYZ?

This article provides a good overview of how the various protocols fit together:

A2A, MCP, AG-UI, A2UI: The Essential 2026 AI Agent Protocol Stack

Confused yet? But wait- there’s more!

How Does A2UI Compare? - A2UI

Antigravity CLI

Antigravity CLI is the follow-on successor to Gemini CLI- the terminal driven, agent assisted coding tool.

Full details on installing Antigravity CLI are here:

Getting Started with Antigravity CLI

Testing the Antigravity CLI Environment

Once you have all the tools in place- you can test the startup of Antigravity CLI.

You will need to authenticate with a Google Cloud Project or your Google Account:

agy

This will start the interface:

Checking the Developer Environment

Verify that all the prerequisite packages and compilers are installed — and clone the sample Github repo:

git clone https://github.com/xbill9/currency-agent
cd currency-agent

Once you have your Google Cloud Project and preferred authentication method — run the init.sh script to validate the setup:

xbill@penguin:~/currency-agent$ source init.sh

[environment: Development]
Updated property [core/project].
ADC is valid.
Environment setup
GOOGLE_GENAI_USE_VERTEXAI=1
GOOGLE_CLOUD_PROJECT=comglitn
GOOGLE_CLOUD_LOCATION=us-central1
IMAGEN_MODEL="imagen-3.0-fast-generate-001"
GENAI_MODEL="gemini-2.5-flash"
Cloud Login
  Credentialed Accounts
ACTIVE ACCOUNT
* [email protected]

The set_env.sh script is provided to set common ADK environment variables:

xbill@penguin:~/currency-agent$ source set_env.sh 
Current Environment
GOOGLE_GENAI_USE_VERTEXAI=false
GOOGLE_CLOUD_PROJECT=comglitn
GOOGLE_CLOUD_LOCATION=us-central1
GENAI_MODEL="gemini-2.5-flash"

Cloud Login
  Credentialed Accounts
ACTIVE ACCOUNT
* [email protected]

To set the active account, run:
    $ gcloud config set account `ACCOUNT`

ADK Version
adk, version 2.1.0

Debugging API Permission Errors

If your application default credentials expires or your Google Cloud Authentication expires you will get an error. The workaround is to re-authenticate:

gcloud auth login
gcloud auth application-default login

Another common error is that the environment variables are not set correctly. Go the the root directory and re-run the set_env.sh to set the variables:

cd ~/currency-agent
source set_env.sh

A2A Debugging Tools -A2A Inspector

The A2A Inspector is a standalone tool that provides low level visibility into the A2A protocol. The GitHub is available here:

GitHub - a2aproject/a2a-inspector: Validation Tools for A2A Agents

A summary of the features of the A2A inspector can be found here:

A2A Protocol Documentation

To install the A2A Inspector:

cd ~
git clone https://github.com/a2aproject/a2a-inspector

Then follow the build instructions — you need uv , and a recent version of node :

a2a-inspector/README.md at main · a2aproject/a2a-inspector

Staring the A2A Inspector

Once the A2A inspector has been installed- you can validate the installation by using this URL:

http://127.0.0.1:5001/

Local A2A Validation with A2A Inspector

The local A2A agent is checked with the A2A Inspector:

> which port is the a2a server listening on

● ListDir(/home/xbill/currency-agent) 
● Read(/home/xbill/currency-agent/GEMINI.md) 
● Read(/home/xbill/currency-agent/Makefile) (ctrl+o to expand)

  The A2A server is listening on port 10000 .

  This is configured when starting the service (e.g., via uvicorn currency_agent.agent:a2a_app --port 10000 ).

and then the Agent card is retrieved:

and the Agent skill cross checked:

Antigravity CLI with MCP Connection

The Master Agent starts a FastMCP server interface. This allows Antigravity CLI to be used as a MCP client:

xbill@penguin:~/currency-agent/.agents$ more mcp_config.json 

    "currency_mcp": {
      "serverUrl": "http://localhost:8080/mcp"
    }

The MCP tools can be reviewed:

MCP Servers

Plugins (~/.gemini/antigravity-cli/plugins)
   ✓ currency_mcp Tools: get_exchange_rate

The MCP tool functionality can be checked with Antigravity:

> what does get_exchange_rate do?

  I will search the repository to locate the definition of the get_exchange_rate tool.

● Search(Grep search) (ctrl+o to expand)

  I will view the definition of the get_exchange_rate tool in server.py.

● Read(/home/xbill/currency-agent/mcp-server/server.py) (ctrl+o to expand)

  The server.py tool is an MCP tool defined in server.py.

  ### What it does:

  It retrieves the currency exchange rate between two currencies for a specific date (or the latest date) using the
  free Frankfurter API https://www.frankfurter.dev/.

  ### Parameters:

  • currency_from ( str , default "USD" ): The source currency code (e.g. "USD" , "GBP" , "CAD" ).
  • currency_to ( str , default "EUR" ): The destination currency code to convert to (e.g. "EUR" , "JPY" , "AUD" 
  ).
  • currency_date ( str , default "latest" ): The date for the exchange rate query (formatted as YYYY-MM-DD or    
  "latest" ).

  ### Returns:

  It returns a Python dictionary containing the Frankfurter API response:

    {
      "amount": 1.0,
      "base": "USD",       
      "date": "2026-06-03",
      "rates": {      
        "EUR": 0.86103
      } 
    }

  If an HTTP request fails or returns an invalid response, it catches the error and returns a dictionary detailing the
  error (e.g., {"error": "API request failed: ..."} ).

So What is all this Doing?

All of the setup and configuration allows you to start debugging and visualizing Agent flows using the A2A protocol. The currency Agent has been validated using the A2A Inspector and the MCP interface via Antigravity CLI.

The make help provides the main project targets:

xbill@penguin:~/currency-agent$ make help
Available commands:
  install - Install all project dependencies (including frontend and react UI)
  start - Start all services in background (MCP + Agent)
  stop - Stop all background services
  status - Check status of background services
  mcp - Start the MCP Server (foreground)
  agent - Start the A2A Agent Server (foreground)
  frontend - Build and start the FastAPI + Vanilla TS frontend server (port 8000)
  react-install - Install dependencies for React + CopilotKit UI
  react-ui - Start React Frontend UI (port 3000)
  react-agent - Start React Frontend Agent (port 8000)
  test-client - Run the A2A Client (test queries)
  e2e-test - Run end-to-end tests (alias for test-client)
  adktest - Run interactive ADK CLI for the agent
  test - Run all tests (pytest)
  frontend-test - Run frontend specific tests
  lint - Run linting checks (ruff)
  format - Auto-format code (ruff)
  clean - Remove caches and logs
  deploy - Deploy to Cloud Run using Cloud Build
  logs - Read logs from Cloud Run
  endpoint - Get the Cloud Run service endpoint
  remote-status - Check the status of the remote endpoint
xbill@penguin:~/currency-agent$ 

Building and Debugging

The Makefile provides targets to build and manage the project:

make install
xbill@penguin:~/currency-agent$ make install
Installing dependencies...
uv sync
Resolved 109 packages in 1ms
Checked 105 packages in 0.60ms
make frontend-install
make[1]: Entering directory '/home/xbill/currency-agent'
Installing frontend dependencies...
cd frontend/frontend && npm install

up to date, audited 16 packages in 606ms

5 packages are looking for funding
  run `npm fund` for details

1 moderate severity vulnerability

To address all issues, run:
  npm audit fix

Run `npm audit` for details.
uv pip install -r frontend/requirements.txt
Checked 14 packages in 7ms
make[1]: Leaving directory '/home/xbill/currency-agent'
make react-install
make[1]: Entering directory '/home/xbill/currency-agent'
Installing React frontend dependencies...
cd frontend-react && npm install

> [email protected] postinstall
> npm run install:agent

> [email protected] install:agent
> ./scripts/setup-agent.sh || scripts\setup-agent.bat

Resolved 127 packages in 0.53ms
Checked 123 packages in 0.56ms

up to date, audited 1195 packages in 2s

232 packages are looking for funding
  run `npm fund` for details

9 vulnerabilities (5 low, 3 moderate, 1 high)

To address issues that do not require attention, run:
  npm audit fix

To address all issues (including breaking changes), run:
  npm audit fix --force

Run `npm audit` for details.
make[1]: Leaving directory '/home/xbill/currency-agent'

Lint:

xbill@penguin:~/currency-agent$ make lint
Running linting checks (ruff check + format)...
uv run ruff check .
All checks passed!
uv run ruff format --check .
20 files already formatted

Test:

xbill@penguin:~/currency-agent$ make test
Running tests...
make stop
make[1]: Entering directory '/home/xbill/currency-agent'
Stopping servers...
make[1]: Leaving directory '/home/xbill/currency-agent'
make start
make[1]: Entering directory '/home/xbill/currency-agent'
Starting MCP Server in background...
Waiting for MCP Server to initialize...
Starting A2A Agent Server in background...
Services started. Logs: mcp.log, agent.log
make[1]: Leaving directory '/home/xbill/currency-agent'
uv run pytest
================================================== test session starts ===================================================
platform linux -- Python 3.13.13, pytest-9.0.3, pluggy-1.6.0
rootdir: /home/xbill/currency-agent
configfile: pyproject.toml
plugins: asyncio-1.3.0, anyio-4.10.0
asyncio: mode=Mode.STRICT, debug=False, asyncio_default_fixture_loop_scope=None, asyncio_default_test_loop_scope=function
collected 15 items                                                                                                       

frontend/tests/test_a2a_utils.py .. [13%]
frontend/tests/test_api.py ... [33%]
frontend/tests/test_auth.py ... [53%]
frontend/tests/test_logging_config.py .. [66%]
frontend/tests/test_utils.py ... [86%]
frontend-react/scripts/test_copilot_endpoint.py s [93%]
mcp-server/test_server.py . [100%]

Time to Start some Currency Arbitrage!

The servers are started:

xbill@penguin:~/currency-agent$ make start
Starting MCP Server in background...
Waiting for MCP Server to initialize...
Starting A2A Agent Server in background...
Services started. Logs: mcp.log, agent.log
xbill@penguin:~/currency-agent$ make status
Checking status of background services...
  All backend and frontend services are currently active and running:

  • MCP Server: Running (Port 8080)
  • A2A Agent Server: Running (Port 10000)
  • Frontend Server: Running (Port 8000)

and checked end-to-end:

xbill@penguin:~/currency-agent$ make e2e-test
Running A2A Client tests...
uv run currency_agent/test_client.py
/home/xbill/currency-agent/.venv/lib/python3.13/site-packages/google/protobuf/runtime_version.py:98: UserWarning: Protobuf gencode version 5.29.3 is exactly one major version older than the runtime version 6.31.1 at a2a.proto. Please update the gencode to avoid compatibility violations in the next runtime release.
  warnings.warn(
--- 🔄 Connecting to agent at http://127.0.0.1:10000... ---
/home/xbill/currency-agent/currency_agent/test_client.py:136: DeprecationWarning: A2AClient is deprecated and will be removed in a future version. Use ClientFactory to create a client with a JSON-RPC transport.
  client = A2AClient(
--- ✅ Connection successful. ---
--- ✉️ Single Turn Request ---
--- 📥 Single Turn Request Response ---
{"id":"0921454b-23dc-4b11-99b0-8e505ce61df9","jsonrpc":"2.0","result":{"artifacts":[{"artifactId":"7d9d4434-d5af-4741-b191-05d939f1c049","parts":[{"data":{"version":"v0.9","updateComponents":{"surfaceId":"currency_agent","components":[{"id":"root","component":"Card","child":"conversion_display"},{"id":"conversion_display","component":"Column","children":["initial_amount","exchange_rate_text","converted_amount"]},{"id":"initial_amount","component":"Text","variant":"h3","text":"100 USD is equal to:"},{"id":"exchange_rate_text","component":"Text","variant":"body","text":"Exchange Rate (USD to CAD): 1 USD = 1.3857 CAD"},{"id":"converted_amount","component":"Text","variant":"h2","text":"138.57 CAD"}]}},"kind":"data","metadata":{"mimeType":"application/json+a2ui"}},{"kind":"text","text":"100 USD is 138.57 CAD."}]}],"contextId":"81d471d1-0a12-4470-b752-85904a0d8d6c","history":[{"contextId":"81d471d1-0a12-4470-b752-85904a0d8d6c","kind":"message","messageId":"be48a82bd4d34484983ed106cc7d3f33","parts":[{"kind":"text","text":"how much is 100 USD in CAD?"}],"role":"user","taskId":"52014164-6d93-430f-9ea6-ecbeea73ac62"},{"contextId":"81d471d1-0a12-4470-b752-85904a0d8d6c","kind":"message","messageId":"be48a82bd4d34484983ed106cc7d3f33","parts":[{"kind":"text","text":"how much is 100 USD in CAD?"}],"role":"user","taskId":"52014164-6d93-430f-9ea6-ecbeea73ac62"},{"kind":"message","messageId":"b9789b81-b159-4f2d-83c5-c7c2b1c90b2b","parts":[{"data":{"id":"adk-c26ec1aa-91f7-446e-8630-252c9d6c906d","args":{"currency_from":"USD","currency_to":"CAD"},"name":"get_exchange_rate"},"kind":"data","metadata":{"adk_type":"function_call","adk_thought_signature":"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"}}],"role":"agent"},{"kind":"message","messageId":"127d8a47-ca80-459e-933e-5d946d33da6a","parts":[{"data":{"id":"adk-c26ec1aa-91f7-446e-8630-252c9d6c906d","name":"get_exchange_rate","response":{"content":[{"type":"text","text":"{\"amount\":1.0,\"base\":\"USD\",\"date\":\"2026-06-03\",\"rates\":{\"CAD\":1.3857}}"}],"structuredContent":{"amount":1.0,"base":"USD","date":"2026-06-03","rates":{"CAD":1.3857}},"isError":false}},"kind":"data","metadata":{"adk_type":"function_response"}}],"role":"agent"},{"kind":"message","messageId":"f8313128-811e-43c1-9ad5-2463bc7ecd61","parts":[{"data":{"version":"v0.9","updateComponents":{"surfaceId":"currency_agent","components":[{"id":"root","component":"Card","child":"conversion_display"},{"id":"conversion_display","component":"Column","children":["initial_amount","exchange_rate_text","converted_amount"]},{"id":"initial_amount","component":"Text","variant":"h3","text":"100 USD is equal to:"},{"id":"exchange_rate_text","component":"Text","variant":"body","text":"Exchange Rate (USD to CAD): 1 USD = 1.3857 CAD"},{"id":"converted_amount","component":"Text","variant":"h2","text":"138.57 CAD"}]}},"kind":"data","metadata":{"mimeType":"application/json+a2ui"}},{"kind":"text","text":"100 USD is 138.57 CAD."}],"role":"agent"}],"id":"52014164-6d93-430f-9ea6-ecbeea73ac62","kind":"task","metadata":{"adk_app_name":"currency_agent","adk_user_id":"A2A_USER_81d471d1-0a12-4470-b752-85904a0d8d6c","adk_session_id":"81d471d1-0a12-4470-b752-85904a0d8d6c","adk_invocation_id":"e-800169b7-70f8-4a90-9613-c80d43152b4f","adk_author":"currency_agent","adk_event_id":"0e81d6c1-5b69-4969-bad7-6d14fe58e9ab","adk_usage_metadata":{"cacheTokensDetails":[{"modality":"TEXT","tokenCount":24060}],"cachedContentTokenCount":24060,"candidatesTokenCount":317,"promptTokenCount":24450,"promptTokensDetails":[{"modality":"TEXT","tokenCount":24450}],"thoughtsTokenCount":100,"totalTokenCount":24867},"adk_actions":{"stateDelta":{},"artifactDelta":{},"requestedAuthConfigs":{},"requestedToolConfirmations":{}}},"status":{"state":"completed","timestamp":"2026-06-03T19:31:49.955541+00:00"}}}

--- ❔ Query Task ---
--- 📥 Query Task Response ---
{"id":"416f7448-0059-45aa-90c6-c90a96d021bc","jsonrpc":"2.0","result":{"artifacts":[{"artifactId":"7d9d4434-d5af-4741-b191-05d939f1c049","parts":[{"data":{"version":"v0.9","updateComponents":{"surfaceId":"currency_agent","components":[{"id":"root","component":"Card","child":"conversion_display"},{"id":"conversion_display","component":"Column","children":["initial_amount","exchange_rate_text","converted_amount"]},{"id":"initial_amount","component":"Text","variant":"h3","text":"100 USD is equal to:"},{"id":"exchange_rate_text","component":"Text","variant":"body","text":"Exchange Rate (USD to CAD): 1 USD = 1.3857 CAD"},{"id":"converted_amount","component":"Text","variant":"h2","text":"138.57 CAD"}]}},"kind":"data","metadata":{"mimeType":"application/json+a2ui"}},{"kind":"text","text":"100 USD is 138.57 CAD."}]}],"contextId":"81d471d1-0a12-4470-b752-85904a0d8d6c","history":[{"contextId":"81d471d1-0a12-4470-b752-85904a0d8d6c","kind":"message","messageId":"be48a82bd4d34484983ed106cc7d3f33","parts":[{"kind":"text","text":"how much is 100 USD in CAD?"}],"role":"user","taskId":"52014164-6d93-430f-9ea6-ecbeea73ac62"},{"contextId":"81d471d1-0a12-4470-b752-85904a0d8d6c","kind":"message","messageId":"be48a82bd4d34484983ed106cc7d3f33","parts":[{"kind":"text","text":"how much is 100 USD in CAD?"}],"role":"user","taskId":"52014164-6d93-430f-9ea6-ecbeea73ac62"},{"kind":"message","messageId":"b9789b81-b159-4f2d-83c5-c7c2b1c90b2b","parts":[{"data":{"id":"adk-c26ec1aa-91f7-446e-8630-252c9d6c906d","args":{"currency_from":"USD","currency_to":"CAD"},"name":"get_exchange_rate"},"kind":"data","metadata":{"adk_type":"function_call","adk_thought_signature":"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"}}],"role":"agent"},{"kind":"message","messageId":"127d8a47-ca80-459e-933e-5d946d33da6a","parts":[{"data":{"id":"adk-c26ec1aa-91f7-446e-8630-252c9d6c906d","name":"get_exchange_rate","response":{"content":[{"type":"text","text":"{\"amount\":1.0,\"base\":\"USD\",\"date\":\"2026-06-03\",\"rates\":{\"CAD\":1.3857}}"}],"structuredContent":{"amount":1.0,"base":"USD","date":"2026-06-03","rates":{"CAD":1.3857}},"isError":false}},"kind":"data","metadata":{"adk_type":"function_response"}}],"role":"agent"},{"kind":"message","messageId":"f8313128-811e-43c1-9ad5-2463bc7ecd61","parts":[{"data":{"version":"v0.9","updateComponents":{"surfaceId":"currency_agent","components":[{"id":"root","component":"Card","child":"conversion_display"},{"id":"conversion_display","component":"Column","children":["initial_amount","exchange_rate_text","converted_amount"]},{"id":"initial_amount","component":"Text","variant":"h3","text":"100 USD is equal to:"},{"id":"exchange_rate_text","component":"Text","variant":"body","text":"Exchange Rate (USD to CAD): 1 USD = 1.3857 CAD"},{"id":"converted_amount","component":"Text","variant":"h2","text":"138.57 CAD"}]}},"kind":"data","metadata":{"mimeType":"application/json+a2ui"}},{"kind":"text","text":"100 USD is 138.57 CAD."}],"role":"agent"}],"id":"52014164-6d93-430f-9ea6-ecbeea73ac62","kind":"task","metadata":{"adk_app_name":"currency_agent","adk_user_id":"A2A_USER_81d471d1-0a12-4470-b752-85904a0d8d6c","adk_session_id":"81d471d1-0a12-4470-b752-85904a0d8d6c","adk_invocation_id":"e-800169b7-70f8-4a90-9613-c80d43152b4f","adk_author":"currency_agent","adk_event_id":"0e81d6c1-5b69-4969-bad7-6d14fe58e9ab","adk_usage_metadata":{"cacheTokensDetails":[{"modality":"TEXT","tokenCount":24060}],"cachedContentTokenCount":24060,"candidatesTokenCount":317,"promptTokenCount":24450,"promptTokensDetails":[{"modality":"TEXT","tokenCount":24450}],"thoughtsTokenCount":100,"totalTokenCount":24867},"adk_actions":{"stateDelta":{},"artifactDelta":{},"requestedAuthConfigs":{},"requestedToolConfirmations":{}}},"status":{"state":"completed","timestamp":"2026-06-03T19:31:49.955541+00:00"}}}

--- 📝 Multi-Turn Request ---
--- 📥 Multi-Turn: First Turn Response ---
{"id":"c17d3d20-d4c1-4b70-b2a3-1f56b30529f1","jsonrpc":"2.0","result":{"artifacts":[{"artifactId":"29c35c34-b8ea-4c29-9ce1-929b97ab3ca3","parts":[{"kind":"text","text":"What currency do you want to convert to?"}]}],"contextId":"a106129a-7692-4bd1-bd06-0e4f09a258af","history":[{"contextId":"a106129a-7692-4bd1-bd06-0e4f09a258af","kind":"message","messageId":"2b709153e93d4ca2b59349c4dba6f658","parts":[{"kind":"text","text":"how much is 100 USD?"}],"role":"user","taskId":"e7eb6014-f770-4b06-b7fb-1a1a12b3bfee"},{"contextId":"a106129a-7692-4bd1-bd06-0e4f09a258af","kind":"message","messageId":"2b709153e93d4ca2b59349c4dba6f658","parts":[{"kind":"text","text":"how much is 100 USD?"}],"role":"user","taskId":"e7eb6014-f770-4b06-b7fb-1a1a12b3bfee"},{"kind":"message","messageId":"45a99a3b-7395-4af5-af7a-c1452e1f53b1","parts":[{"kind":"text","text":"What currency do you want to convert to?"}],"role":"agent"}],"id":"e7eb6014-f770-4b06-b7fb-1a1a12b3bfee","kind":"task","metadata":{"adk_app_name":"currency_agent","adk_user_id":"A2A_USER_a106129a-7692-4bd1-bd06-0e4f09a258af","adk_session_id":"a106129a-7692-4bd1-bd06-0e4f09a258af","adk_invocation_id":"e-8041dafe-4319-42ed-89ef-0999dacf84a7","adk_author":"currency_agent","adk_event_id":"195fc334-df88-40ec-bf75-884ff395b671","adk_usage_metadata":{"candidatesTokenCount":9,"promptTokenCount":24309,"promptTokensDetails":[{"modality":"TEXT","tokenCount":24309}],"thoughtsTokenCount":61,"totalTokenCount":24379},"adk_actions":{"stateDelta":{},"artifactDelta":{},"requestedAuthConfigs":{},"requestedToolConfirmations":{}}},"status":{"state":"completed","timestamp":"2026-06-03T19:31:51.471602+00:00"}}}

--- 🚀 First turn completed, no further input required for this test case. ---

Start the Currency Session

Once all the agents have been started — the web interface on port 8000 is connected with a standard web browser:

The new A2UI tools are presented:

Running the new A2UI tools — conversion card:

Rates Table:

Line Chart: 5 day trend:

Multicurrency Wallet:

What does the A2UI Code Look Like?

The A2UI generated code follows a well defined format:

 To match the workspace's glassmorphic aesthetics, the agent typically wraps components inside a Card layout using  
  parentId relationships:

    {                                                                                                                  
      "version": "v0.9",
      "components": [
        {
          "id": "container_card",
          "type": "Card",
          "props": {
            "title": "📈 Historical exchange rates"
          }
        },
        {
          "id": "historical_line_chart",
          "type": "LineChart",
          "parentId": "container_card",
          "props": {
            "title": "USD / EUR Trend",
            "labels": ["Mon", "Tue", "Wed", "Thu", "Fri"],
            "values": [0.8558, 0.8577, 0.8573, 0.8583, 0.8610],
            "color": "#3b82f6"
          }
        }
      ]
    }

How does this all Work?

Antigravity CLI has the answer!

> summarize how the currency project uses A2UI

▸ Thought for 3s, 422 tokens
  Analyzing A2UI Integration
  The Currency Agent uses Agent-to-UI (A2UI) to present structured information (like conversion rates, trend tables,
  and
  wallet balances) as interactive, premium UI cards instead of raw text. Here is how it is implemented across the
  project:

  ### 1. Agent Configuration & Generation (agent.py)

  • Schema Definition: Initializes agent.py with version 0.9 components (e.g., Card , Text , Table , Column ,
  Row ).
  • Prompt Injection: The schema manager generates a detailed system prompt instruction that details the catalog schema
  and provides examples. This instructs the LLM (agent.py) on how to output structured JSON wrapped inside <a2ui-
  json>...</a2ui-json> blocks.
  • Part Conversion: Integrates agent.py inside the A2A executor factory to process Gemini's output and format
  A2UI blocks during Agent-to-Agent (A2A) protocol exchanges.

  ### 2. Frontend Parser & Renderer (app.ts)

  • Response Parsing: The workspace frontend regex-searches all incoming chat messages for <a2ui-json> blocks.
  • Dynamic Rendering: Passes the JSON content to app.ts, which parses the tree of UI components and
  dynamically builds native HTML elements matching the catalog parameters.
  • Sandbox Demos: Implements sandbox controls that simulate agent responses to let developers preview A2UI components
  live in the web browser.

  ### 3. Visual Styling (style.css)

  • Defines CSS properties for .a2ui-card and .a2ui-block-container , styled with glassmorphic borders, neon mint
  green text, and premium dark-mode grids matching the theme.

Phew. that's a lot! What else is there?

  ### How A2UI works in this Project

  Rather than running a bidirectional state synchronizer (like AG-UI), this project relies on a clean, declarative A2UI
  flow:

  1. Prompt Configuration: In agent.py, the agent is configured with the A2uiSchemaManager to append formatting
  rules and examples to the system prompt.
  2. Declarative Output: The LLM output contains UI components formatted in JSON, wrapped in <a2ui-json>...</a2ui-    
  json> blocks.
  3. A2A Integration: The agent.py extracts these blocks and converts them into A2A parts.
  4. Client-Side Rendering: The frontend browser client (app.ts) extracts the raw <a2ui-json> blocks via
  regex and calls its client-side rendering engine app.ts to map the JSON elements to styled HTML containers.

Are you sure? How about a code review:

  Initiating Code Review
  Here is a detailed code review of how Agent-to-UI (A2UI) is used within this project.
  ──────
   A2UI Code Review: Currency Agent Ecosystem 

  ## 1. Backend Architecture & Integration (agent.py)

  The backend uses the Google ADK library and the official a2ui Python package to inject formatting rules into the
  model context and package UI components.

  ### Setup & Prompt Generation
  • Schema Definition: Initializes the A2uiSchemaManager with version 0.9 and selects the BasicCatalog .
  • System Prompt Injection:
    SYSTEM_INSTRUCTION = schema_manager.generate_system_prompt(                                                        
        role_description=( ... ),                                                                                      
        workflow_description="Use components to display structured currency conversions and rates.",                   
        ui_description="Use Card and Text components for displaying rates. Use Table components when presenting        
  multiple exchange rates or history.",                                                                                
        include_schema=True,                                                                                           
        include_examples=True,                                                                                         
        allowed_components=["Card", "Text", "Table"],                                                                  
    )                                                                                                                  

      • Review Verdict: Excellent practice. Rather than manually typing and maintaining JSON-schema rules or few-shot
      examples in a raw text string, the A2uiSchemaManager generates precise prompt instructions dynamically. Setting
      allowed_components limits the model's vocabulary and prevents hallucination of unsupported UI elements.

  ### Toolset & Part Conversion

  • UI Toolset: Equips the LlmAgent with a2ui_toolset ( SendA2uiToClientToolset ).
  • Part Converter Setup:
    def agent_executor_factory(runner):                                                                                
        config = A2aAgentExecutorConfig(                                                                               
            gen_ai_part_converter=A2uiPartConverter(                                                                   
                selected_catalog, bypass_tool_check=True                                                               
            ).convert                                                                                                  
        )                                                                                                              
        return A2aAgentExecutor(runner=runner, config=config)                                                          

      • Review Verdict: By integrating the A2uiPartConverter directly into the A2aAgentExecutorConfig , any text
      responses returned by Gemini containing <a2ui-json> blocks are interceptable and mapped cleanly into structured
      data parts matching the A2A MIME standard ( application/json+a2ui ).

  ──────
  ## 2. Frontend Parser & Renderer (app.ts)

  The frontend receives the response, parses out the <a2ui-json> block, and recursively translates it to styled HTML
  components.

  ### Regex Extraction

  • Location: app.ts
    const a2uiRegex = /<a2ui-json>([\s\S]*?)<\/a2ui-json>/;                                                            
    const match = text.match(a2uiRegex);                                                                               

      • Review Verdict: The regex is simple, clean, and handles multi-line blocks. It strips the <a2ui-json> block
      from the text message so the user gets clean text alongside a structured component.

  ### DOM Generation Engine ( renderA2UI )

  • Location: app.ts
  • Schema Normalization: Parses the JSON and accommodates multiple layout formats (basic array, parsed.components ,  
  parsed.updateComponents.components , or parsed.message.components ).
  • Hierarchy Resolution: Resolves hierarchical relationships using parentId referencing, allowing nested elements:
    normalized.forEach((c: any) => {                                                                                   
        if (c.child) { ... }                                                                                           
        if (Array.isArray(c.children)) { ... }                                                                         
    });                                                                                                                

  • Security & XSS Prevention:
  Throughout the element building code (e.g. el.textContent = c.props.value or title.textContent = cardComp.props.  
  title ), the renderer uses textContent rather than innerHTML .
      • Review Verdict: Highly Secure. Directly mapping LLM output to innerHTML poses a serious Cross-Site Scripting
      (XSS) risk if the model produces malformed or malicious script tags. Using textContent ensures that strings are
      treated strictly as text.

  ### Visual Aesthetics

  • The components are styled using theme-consistent CSS variables (e.g., background = 'hsl(var(--bg-card))' , border
  = '1px solid hsl(var(--border-color))' ).
      • Review Verdict: Ensures that the A2UI elements match the surrounding dark mode/glassmorphism styling system
      seamlessly.

  ──────
  ## 3. Strengths & Best Practices Found

  1. Decoupled Architecture: The agent provides a declarative specification of the UI components ( what to render) and
  leaves the client browser in control of styling, positioning, and visual mapping ( how to render).
  2. Robust Fallbacks: The code handles cases where there is a parse failure by logging the error and falling back to
  raw text so that the conversation does not break.
  3. Recursive Column/Row Layouts: Incorporating parent-child rendering ( renderElement ) allows complex nested grids
  (e.g., wallet columns/rows) to render gracefully.

Summary

The goal of the demo/article was to enable A2UI presentation controls on a functional agent supporting common protocols like A2A and MCP.

The currency agent was chosen as a starting point and new A2UI presentation objects were added to the user interface with Antigravity CLI.

Finally- the updated A2UI project integration was summarized, and code reviewed with Antigravity CLI.