AWS Builder Center
Google ADK Cross Cloud to Amazon Bedrock over A2A

Google ADK Cross Cloud to Amazon Bedrock over A2A

For true cross cloud A2A coverage- the path must be traced both ways. This article flips the script and traces the ADK to Bedrock case.

Making the Impossible - Possible!


This is the third run of the same cross-cloud currency benchmark, and the first with the arrow pointing the other way. A Google ADK master on Cloud Run (Gemini 2.5 Flash,  us-central1 ) owns the benchmark policy. It calls an Amazon Bedrock AgentCore worker (Strands Agents on Nova Micro,  us-east-1 ) over A2A v1.0, and cross-checks the answer against an MCP exchange-rate tool.
The previous run had Bedrock as the master and ADK as the worker. Reversing it was supposed to be a deployment change — the domain core is framework-independent, so the comparison logic never moved. What actually happened is that reversal exposed six interoperability defects, five of which no local test could have caught. That is the interesting part, so this article leads with the failures.

What is this project trying to do?

Most A2A demos stop at "the HTTP 200 came back." That is a smoke test, not an interoperability benchmark.
The benchmark runs three modes so the cost of remote verification is separable from the cost of the work:
Mode
What runs
Purpose
 mcp_only 
ADK master calls the rate tool over MCP stdio
Baseline
 a2a_only 
ADK master delegates to the AgentCore worker
Remote-agent cost
 verified 
Both, concurrently, then compared deterministically
Accuracy/overhead
Comparison is arithmetic, not a model judgement. Amounts and rates are  Decimal . A model is never asked whether two numbers agree.



Six things that broke

1. The Strands A2A extra speaks the wrong protocol version

This one was caught before deploying, and it dictated the whole worker design.
 strands-agents[a2a]  pins  a2a-sdk<0.4  — the A2A v0.3 wire methods ( message/send ).  google-adk[a2a]==2.5.0  uses  a2a-sdk  1.x — v1.0 ( SendMessage ). A v1.0 client cannot call a v0.3 server, and the agent card carries no version negotiation.
This is the same split that forced the A2UI extension out of the ADK agent in the first run of this project. It reappeared from the opposite side.
The fix: use  strands-agents  for the agent loop only, and build the A2A v1.0 server surface from  a2a-sdk  directly.
# app/CurrencyWorker/main.py -- NOT strands.multiagent.a2a.A2AServer
from a2a.server.request_handlers import DefaultRequestHandler
from a2a.server.routes import create_agent_card_routes, create_jsonrpc_routes

app = Starlette(routes=[
*create_agent_card_routes(AGENT_CARD),
*create_jsonrpc_routes(_request_handler, rpc_url="/"),
])
A test asserts the lockfile keeps  a2a-sdk  on 1.x, so a dependency bump cannot quietly reintroduce the mismatch.

2.  google-adk[a2a]  does not install its own server dependency

The Cloud Run container built fine, passed every local test, and then died on startup:
ModuleNotFoundError: No module named 'sse_starlette'
File ".../google/adk/a2a/_compat.py", line 769, in attach_a2a_routes_to_app
ADK's  to_a2a()  imports  a2a.server.routes , which needs  sse_starlette , which arrives only with the  a2a-sdk[http-server]  extra. Neither  google-adk[a2a]  nor bare  a2a-sdk  pulls it in.
Cost: one failed rollout. Both halves now declare  a2a-sdk[http-server] .

3. The IAM trust policy could never have matched

The coordinator holds no AWS keys. It federates: Cloud Run's metadata server mints a Google OIDC token, STS  AssumeRoleWithWebIdentity  exchanges it for temporary credentials, and requests are SigV4-signed.
The trust policy looked obviously correct and was silently impossible:
"Condition": { "StringEquals": {
"accounts.google.com:aud": "currencybench-agentcore-worker",
"accounts.google.com:sub": "1019138736740282766.."
}}
AWS does not map those condition keys to the claims their names suggest:
Condition key
Actual Google claim
 accounts.google.com:oaud 
the token's  aud 
 accounts.google.com:aud 
the token's  azp  (numeric client id)
 accounts.google.com:sub 
the token's  sub 
The audience string was being compared against a number. Pin the audience with  oaud , not  aud .

4. Creating an OIDC provider for Google breaks Google federation

The natural next step — register  accounts.google.com  as an IAM OIDC identity provider — is wrong. AWS federates with Google natively. Creating an explicit provider switched STS to validating against that provider's thumbprint, and every exchange failed:
<Code>InvalidIdentityToken</Code>
<Message>The web identity token provided could not be validated.</Message>
Deleting the provider fixed it. The principal is the bare domain:
"Principal": { "Federated": "accounts.google.com" }
Worth noting the failure mode: an invalid token and a rejected trust policy are different errors ( InvalidIdentityToken  vs  AccessDenied ), and that distinction is the fastest way to tell these two bugs apart.

5. AgentCore strips the A2A version header, and the SDK defaults to v0.3

With federation working and SigV4 signing accepted, the worker rejected its own correctly-versioned client:
A2A version '0.3' is not supported by this handler. Expected version '1.0'.
The client was  a2a-sdk  1.1.2 and does send the right header ( client_factory.py  sets  A2A-Version: 1.0 ). The worker was v1.0. The card advertised  "protocolVersion": "1.0" .
The header never arrived. AgentCore's A2A runtime forwards only allow-listed request headers — and the server-side validator treats a missing header as v0.3:
# a2a/utils/version_validator.py
if not actual_version:
return constants.PROTOCOL_VERSION_0_3
A dropped header and an old client are indistinguishable to the server. The fix is one line of runtime config:
"requestHeaderAllowlist": ["A2A-Version"]
This is the version-skew theme of the whole project arriving a third way: first through a transitive pin, then through a framework extra, now through a proxy.

6. Agent cards advertise bind addresses, in both directions

The AgentCore worker's card advertises  127.0.0.1:9000  — its container's own bind address. ADK's  to_a2a()  does the same with  127.0.0.1:8080 .  a2a-sdk  clients route by card URL, so cross-cloud calls fail unless the client rewrites the interfaces to the endpoint it actually dialled.
Documented in the first run for ADK; it is not an ADK quirk.

Bonus: the A2A endpoint is not the URL that looks like a base URL

 agentcore status  prints a URL with the runtime ARN percent-encoded into the path and an  /invocations  suffix. That full URL, suffix included, is the A2A base:
https://bedrock-agentcore.us-east-1.amazonaws.com/runtimes/arn%3Aaws%3A...%2F<id>/invocations
└── card at <base>/.well-known/agent-card.json
Dropping  /invocations  — the tidier-looking base — returns  UnknownOperationException . So does using a bare runtime id instead of the encoded ARN. The percent-encoding also breaks naive  ${VAR##*/runtimes/}  shell parsing when deriving the runtime id for a scoped IAM policy.



What it measures

Five warm invocations per mode, driven through the deployed Cloud Run coordinator,  us-central1  →  us-east-1 , on 2026-07-31:
Mode
Median
Min
Max
 mcp_only 
2.96 s
2.37 s
3.84 s
 a2a_only 
18.92 s
16.30 s
19.68 s
 verified 
16.19 s
15.64 s
22.86 s
Two things stand out.
The remote-agent hop costs about 6× the tool baseline. That is not network latency —  us-central1  to  us-east-1  is tens of milliseconds. It is a second model turn: the worker is an agent, so delegation means Nova Micro plans a tool call, calls it, and writes a structured reply, on top of Gemini doing the same. A2A delegation buys independence, and independence costs an inference.
 verified  is not slower than  a2a_only . Verified mode does the MCP call and the A2A call and compares them, yet its median is lower. The coordinator issues both concurrently:
mcp_task = self._call("mcp", self._mcp, request, failures)
a2a_task = self._call("a2a", self._a2a, request, failures)
mcp_quotes, a2a_quotes = await asyncio.gather(mcp_task, a2a_task)
So verified ≈ max(mcp, a2a), not the sum. Independent verification is close to free once you are already paying for the remote agent. The two medians differ by less than the run-to-run spread, so read them as equal, not as verified being genuinely faster.
Correctness, verified mode, 100 USD → EUR and CHF:
EUR rate 0.87138 -> 87.138 CHF rate 0.81248 -> 81.248
primary: mcp-stdio:frankfurter-live
verifier: aws-agentcore-a2a-worker
Both targets agreed. Both also carried  rate is stale; check the observation timestamp  — the MCP side returned Frankfurter's daily reference rate stamped  2026-07-30T00:00:00Z  against a run at  04:09Z  on the 31st, over the 24 h threshold. The staleness rule firing on a real published rate, rather than a fixture, is the more useful signal here.



What this run does not establish

  • The IAM policy is not yet least-privilege. Scoping  bedrock-agentcore:InvokeAgentRuntime  to  runtime/<id>  and  runtime/<id>/*  was denied with 403; only  Resource: "*"  worked. AgentCore authorises against some ARN shape I have not identified, and data-plane events are not in CloudTrail by default. The measurements above were taken with the wide policy.
  • n is 5, one region pair, one day, one model pair. No p95, no cost accounting, no token counts, no cold/warm distribution. The 38-case × 3-mode matrix has not been run in this direction.
  • Nothing here is a claim about which cloud is faster. Gemini 2.5 Flash and Nova Micro are different models at different price points; the 6× is the cost of agent delegation, not of AWS.

The part worth keeping

The domain core did not change.  CurrencyCoordinator  takes two duck-typed adapters, so which cloud hosts the master is a deployment decision. That held up: the comparison logic, the staleness rule, the failure taxonomy, and the evaluation harness all survived the reversal untouched.
Everything that broke was at the seams — protocol versions, dependency extras, identity federation, header forwarding, URL shapes. Five of the six defects were invisible until something real was deployed. If you take one thing from this: an interoperability test that never leaves your laptop is testing your mocks.
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article