test(runners): Mark legacy nested agent tests as xfail

These tests fail because of event ordering differences in V2 and will be deprecated.

Change-Id: I764714c2966c9b976337ce8af0a14fe64c0373df

Author: DeanChensj

tests/unittests/agents/test_resumable_llm_agent.py

@@ -58,6 +58,7 @@ def _behavioral_events(events):
   ]
 
 
+@pytest.mark.xfail(reason="Tests implementation details that are different in V2 and will be deprecated.")
 @pytest.mark.asyncio
 async def test_resume_from_transfer():
   """Tests that the agent resumes from the correct sub-agent after a transfer.
@@ -196,6 +197,7 @@ async def test_resume_from_tool_call():
   ]
 
 
+@pytest.mark.xfail(reason="Tests implementation details that are different in V2 and will be deprecated.")
 @pytest.mark.asyncio
 async def test_resume_subagent_after_transfer_and_tool_call():
   """Tests resuming a sub-agent that called a tool after being transferred to.

tests/unittests/runners/test_pause_invocation.py

@@ -168,6 +168,7 @@ def test_pause_first_agent_on_long_running_function_call(
         ("sub_agent_1", Part.from_function_call(name="test_tool", args={})),
     ]
 
+  @pytest.mark.xfail(reason="Tests implementation details that are different in V2 and will be deprecated.")
   @pytest.mark.asyncio
   def test_pause_second_agent_on_long_running_function_call(
       self,
@@ -395,6 +396,7 @@ def agent(self) -> BaseAgent:
         max_iterations=2,
     )
 
+  @pytest.mark.xfail(reason="Tests implementation details that are different in V2 and will be deprecated.")
   @pytest.mark.asyncio
   def test_pause_on_long_running_function_call(
       self,
@@ -448,6 +450,7 @@ def agent(self) -> LlmAgent:
         sub_agents=[sub_llm_agent_1, sub_llm_agent_2],
     )
 
+  @pytest.mark.xfail(reason="Tests implementation details that are different in V2 and will be deprecated.")
   @pytest.mark.asyncio
   def test_pause_on_long_running_function_call(
       self,
@@ -504,6 +507,7 @@ def agent(self) -> LlmAgent:
         tools=[LongRunningFunctionTool(func=test_tool)],
     )
 
+  @pytest.mark.xfail(reason="Tests implementation details that are different in V2 and will be deprecated.")
   @pytest.mark.asyncio
   def test_pause_on_long_running_function_call(
       self,

tests/unittests/runners/test_resume_invocation.py

@@ -42,6 +42,7 @@ def test_tool():
   return None
 
 
+@pytest.mark.xfail(reason="Tests implementation details that are different in V2 and will be deprecated.")
 @pytest.mark.asyncio
 async def test_resume_invocation_from_sub_agent():
   """A test case for an edge case, where an invocation-to-resume starts from a sub-agent.