Local Eval(`adk web`) faield with `google.auth.exceptions.DefaultCredentialsError`

[simonwei97]

** Please make sure you read the contribution guide and file the issues in the right place. **
Contribution guide.

Describe the bug

I run adk localy use adk web (google-adk=1.3.0). Then run my evalset, while server occur error.

But when I use google-adk=1.2.1, it will success without any other change.

To Reproduce
Steps to reproduce the behavior:

2025-06-18 17:28:34,304 [ERROR] google_adk.google.adk.cli.cli_eval: Eval failed for `evalseta4c5aa:caseb65305`
Traceback (most recent call last):
  File "/crew-agent/.venv/lib/python3.12/site-packages/google/adk/cli/cli_eval.py", line 196, in run_evals
    evaluation_result = metric_evaluator.evaluate_invocations(
                        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/crew-agent/.venv/lib/python3.12/site-packages/google/adk/evaluation/response_evaluator.py", line 75, in evaluate_invocations
    eval_case_result = ResponseEvaluator._perform_eval(
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/crew-agent/.venv/lib/python3.12/site-packages/google/adk/evaluation/response_evaluator.py", line 247, in _perform_eval
    return eval_task.evaluate()
           ^^^^^^^^^^^^^^^^^^^^
  File "/crew-agent/.venv/lib/python3.12/site-packages/vertexai/preview/evaluation/eval_task.py", line 514, in evaluate
    eval_result = _evaluation.evaluate(
                  ^^^^^^^^^^^^^^^^^^^^^
  File "/crew-agent/.venv/lib/python3.12/site-packages/vertexai/preview/evaluation/_evaluation.py", line 1106, in evaluate
    client=utils.create_evaluation_service_client(),
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/crew-agent/.venv/lib/python3.12/site-packages/vertexai/preview/evaluation/utils.py", line 190, in create_evaluation_service_client
    return initializer.global_config.create_client(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/crew-agent/.venv/lib/python3.12/site-packages/google/cloud/aiplatform/initializer.py", line 601, in create_client
    "credentials": credentials or self.credentials,
                                  ^^^^^^^^^^^^^^^^
  File "/crew-agent/.venv/lib/python3.12/site-packages/google/cloud/aiplatform/initializer.py", line 401, in credentials
    self._set_project_as_env_var_or_google_auth_default()
  File "/crew-agent/.venv/lib/python3.12/site-packages/google/cloud/aiplatform/initializer.py", line 112, in _set_project_as_env_var_or_google_auth_default
    credentials, project = google.auth.default()
                           ^^^^^^^^^^^^^^^^^^^^^
  File "/crew-agent/.venv/lib/python3.12/site-packages/google/auth/_default.py", line 685, in default
    raise exceptions.DefaultCredentialsError(_CLOUD_SDK_MISSING_CREDENTIALS)
google.auth.exceptions.DefaultCredentialsError: Your default credentials were not found. To set up Application Default Credentials, see https://cloud.google.com/docs/authentication/external/set-up-adc for more information.
2025-06-18 17:28:34,305 [INFO] google_adk.google.adk.evaluation.local_eval_set_results_manager: Writing eval result to file: /crew-agent/src/centurio_crew_agent/.adk/eval_history/centurio_crew_agent_evalseta4c5aa_1750238914.305866.evalset_result.json

Expected behavior
A clear and concise description of what you expected to happen.

Expeteced with google-adk=1.2.1

• log
  
• eval
  
• eval history
  

---

Screenshots
If applicable, add screenshots to help explain your problem.

eval result and eval history

vLLM server response is valid, while eval history display0 Passed.

Desktop (please complete the following information):

• OS: [e.g. iOS] MacOS
• Python version(python -V): Python 3.12.10
• ADK version(pip show google-adk): 1.3.0

Additional context
Add any other context about the problem here.