Moves unittests to root folder and adds github action to run unit tests. (#72)

* Move unit tests to root package.

* Adds deps to "test" extra, and mark two broken tests in tests/unittests/auth/test_auth_handler.py

* Adds github workflow

* minor fix in lite_llm.py for python 3.9.

* format pyproject.toml

tests/unittests/tools/test_agent_tool.py

@@ -0,0 +1,167 @@
+# Copyright 2025 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from google.adk.agents import Agent
+from google.adk.agents.callback_context import CallbackContext
+from google.adk.tools.agent_tool import AgentTool
+from google.genai.types import Part
+from pydantic import BaseModel
+import pytest
+from pytest import mark
+
+from .. import utils
+
+pytestmark = pytest.mark.skip(
+    reason='Skipping until tool.func evaluations are fixed (async)'
+)
+
+
+function_call_custom = Part.from_function_call(
+    name='tool_agent', args={'custom_input': 'test1'}
+)
+
+function_call_no_schema = Part.from_function_call(
+    name='tool_agent', args={'request': 'test1'}
+)
+
+function_response_custom = Part.from_function_response(
+    name='tool_agent', response={'custom_output': 'response1'}
+)
+
+function_response_no_schema = Part.from_function_response(
+    name='tool_agent', response={'result': 'response1'}
+)
+
+
+def change_state_callback(callback_context: CallbackContext):
+  callback_context.state['state_1'] = 'changed_value'
+  print('change_state_callback: ', callback_context.state)
+
+
+def test_no_schema():
+  mock_model = utils.MockModel.create(
+      responses=[
+          function_call_no_schema,
+          'response1',
+          'response2',
+      ]
+  )
+
+  tool_agent = Agent(
+      name='tool_agent',
+      model=mock_model,
+  )
+
+  root_agent = Agent(
+      name='root_agent',
+      model=mock_model,
+      tools=[AgentTool(agent=tool_agent)],
+  )
+
+  runner = utils.InMemoryRunner(root_agent)
+
+  assert utils.simplify_events(runner.run('test1')) == [
+      ('root_agent', function_call_no_schema),
+      ('root_agent', function_response_no_schema),
+      ('root_agent', 'response2'),
+  ]
+
+
+def test_update_state():
+  """The agent tool can read and change parent state."""
+
+  mock_model = utils.MockModel.create(
+      responses=[
+          function_call_no_schema,
+          '{"custom_output": "response1"}',
+          'response2',
+      ]
+  )
+
+  tool_agent = Agent(
+      name='tool_agent',
+      model=mock_model,
+      instruction='input: {state_1}',
+      before_agent_callback=change_state_callback,
+  )
+
+  root_agent = Agent(
+      name='root_agent',
+      model=mock_model,
+      tools=[AgentTool(agent=tool_agent)],
+  )
+
+  runner = utils.InMemoryRunner(root_agent)
+  runner.session.state['state_1'] = 'state1_value'
+
+  runner.run('test1')
+  assert (
+      'input: changed_value' in mock_model.requests[1].config.system_instruction
+  )
+  assert runner.session.state['state_1'] == 'changed_value'
+
+
+@mark.parametrize(
+    'env_variables',
+    [
+        'GOOGLE_AI',
+        # TODO(wanyif): re-enable after fix.
+        # 'VERTEX',
+    ],
+    indirect=True,
+)
+def test_custom_schema():
+  class CustomInput(BaseModel):
+    custom_input: str
+
+  class CustomOutput(BaseModel):
+    custom_output: str
+
+  mock_model = utils.MockModel.create(
+      responses=[
+          function_call_custom,
+          '{"custom_output": "response1"}',
+          'response2',
+      ]
+  )
+
+  tool_agent = Agent(
+      name='tool_agent',
+      model=mock_model,
+      input_schema=CustomInput,
+      output_schema=CustomOutput,
+      output_key='tool_output',
+  )
+
+  root_agent = Agent(
+      name='root_agent',
+      model=mock_model,
+      tools=[AgentTool(agent=tool_agent)],
+  )
+
+  runner = utils.InMemoryRunner(root_agent)
+  runner.session.state['state_1'] = 'state1_value'
+
+  assert utils.simplify_events(runner.run('test1')) == [
+      ('root_agent', function_call_custom),
+      ('root_agent', function_response_custom),
+      ('root_agent', 'response2'),
+  ]
+
+  assert runner.session.state['tool_output'] == {'custom_output': 'response1'}
+
+  assert len(mock_model.requests) == 3
+  # The second request is the tool agent request.
+  assert mock_model.requests[1].config.response_schema == CustomOutput
+  assert mock_model.requests[1].config.response_mime_type == 'application/json'