get_fast_api_app() fails with `session_db_kwargs` when using sqlite://

[ftnext]

Summary

After upgrading to ADK v1.23.0, get_fast_api_app() fails when session_service_uri uses sqlite:// and session_db_kwargs is provided.

Steps to reproduce

from google.adk.cli.fast_api import get_fast_api_app

app = get_fast_api_app(
    agents_dir="foo",
    session_service_uri="sqlite:///./session.db",
    session_db_kwargs={"pool_size": 10},
    web=False,
)

Actual behavior

Traceback (most recent call last):
  File "/.../google-adk-python/script.py", line 3, in <module>
    app = get_fast_api_app(
        agents_dir="foo",
    ...<2 lines>...
        web=False,
    )
  File "/.../google-adk-python/src/google/adk/cli/fast_api.py", line 122, in get_fast_api_app
    session_service = create_session_service_from_options(
        base_dir=agents_dir,
    ...<2 lines>...
        use_local_storage=use_local_storage,
    )
  File "/.../google-adk-python/src/google/adk/cli/utils/service_factory.py", line 193, in create_session_service_from_options
    service = registry.create_session_service(session_service_uri, **kwargs)
  File "/.../google-adk-python/src/google/adk/cli/service_registry.py", line 132, in create_session_service
    return self._session_factories[scheme](uri, **kwargs)
           ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^
  File "/.../google-adk-python/src/google/adk/cli/service_registry.py", line 255, in sqlite_session_factory
    return SqliteSessionService(db_path=db_path, **kwargs_copy)
TypeError: SqliteSessionService.__init__() got an unexpected keyword argument 'pool_size'

Expected behavior

No error

Environment

• ADK version: 1.23.0
• Python: 3.13.8
• OS: macOS

Additional context

SqliteSessionService only accepts db_path, but the sqlite session factory passes **kwargs.

[wuliang229]

@ftnext you closed this without the fix submitted? Is this intended?

[ftnext]

@wuliang229 I noticed you closed my pull request without comment and adopted the Claude Code commit instead, so I understood this to mean the issue has been resolved.
Should I expect this to be merged to main soon, or is there something else needed?

[wuliang229]

Yeah we will merge that code this week. I was gonna close this when it is merged, but either way works.