test: improve deterministic unit test coverage

tests/agent/test_runner_core.py

@@ -5,12 +5,11 @@ from __future__ import annotations
 
 import asyncio
 import time
-from unittest.mock import AsyncMock, MagicMock
+from unittest.mock import AsyncMock, MagicMock, patch
 
 import pytest
 
 from nanobot.config.schema import AgentDefaults
-from nanobot.agent.tools.registry import ToolRegistry
 from nanobot.providers.base import LLMProvider, LLMResponse, ToolCallRequest
 
 _MAX_TOOL_RESULT_CHARS = AgentDefaults().max_tool_result_chars
@@ -18,7 +17,7 @@ _MAX_TOOL_RESULT_CHARS = AgentDefaults().max_tool_result_chars
 
 @pytest.mark.asyncio
 async def test_runner_preserves_reasoning_fields_and_tool_results():
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     provider = MagicMock(spec=LLMProvider)
     captured_second_call: list[dict] = []
@@ -75,7 +74,7 @@ async def test_runner_preserves_reasoning_fields_and_tool_results():
 
 @pytest.mark.asyncio
 async def test_runner_returns_max_iterations_fallback():
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     provider = MagicMock(spec=LLMProvider)
     provider.chat_with_retry = AsyncMock(return_value=LLMResponse(
@@ -106,7 +105,7 @@ async def test_runner_returns_max_iterations_fallback():
 
 @pytest.mark.asyncio
 async def test_runner_times_out_hung_llm_request():
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     provider = MagicMock(spec=LLMProvider)
 
@@ -136,15 +135,15 @@ async def test_runner_times_out_hung_llm_request():
 @pytest.mark.asyncio
 async def test_runner_does_not_apply_outer_wall_timeout_to_streaming_requests():
     from nanobot.agent.hook import AgentHook, AgentHookContext
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     provider = MagicMock(spec=LLMProvider)
     streamed: list[str] = []
 
     async def chat_stream_with_retry(*, on_content_delta, **kwargs):
-        await asyncio.sleep(0.08)
+        await asyncio.sleep(0)
         await on_content_delta("still ")
-        await asyncio.sleep(0.08)
+        await asyncio.sleep(0)
         await on_content_delta("alive")
         return LLMResponse(content="still alive", tool_calls=[])
 
@@ -161,25 +160,28 @@ async def test_runner_does_not_apply_outer_wall_timeout_to_streaming_requests():
             streamed.append(delta)
 
     runner = AgentRunner(provider)
-    result = await runner.run(AgentRunSpec(
-        initial_messages=[{"role": "user", "content": "think for a while"}],
-        tools=tools,
-        model="test-model",
-        max_iterations=1,
-        max_tool_result_chars=_MAX_TOOL_RESULT_CHARS,
-        hook=StreamingHook(),
-        llm_timeout_s=0.01,
-    ))
+    wait_for = AsyncMock(side_effect=AssertionError("streaming path must not use wait_for"))
+    with patch("nanobot.agent.runner.asyncio.wait_for", wait_for):
+        result = await runner.run(AgentRunSpec(
+            initial_messages=[{"role": "user", "content": "think for a while"}],
+            tools=tools,
+            model="test-model",
+            max_iterations=1,
+            max_tool_result_chars=_MAX_TOOL_RESULT_CHARS,
+            hook=StreamingHook(),
+            llm_timeout_s=0.01,
+        ))
 
     assert result.stop_reason == "completed"
     assert result.final_content == "still alive"
     assert streamed == ["still ", "alive"]
     provider.chat_with_retry.assert_not_awaited()
+    wait_for.assert_not_awaited()
 
 
 @pytest.mark.asyncio
 async def test_runner_replaces_empty_tool_result_with_marker():
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     provider = MagicMock(spec=LLMProvider)
     captured_second_call: list[dict] = []
@@ -218,7 +220,7 @@ async def test_runner_replaces_empty_tool_result_with_marker():
 @pytest.mark.asyncio
 async def test_runner_retries_empty_final_response_with_summary_prompt():
     """Empty responses get 2 silent retries before finalization kicks in."""
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     provider = MagicMock(spec=LLMProvider)
     calls: list[dict] = []
@@ -263,7 +265,7 @@ async def test_runner_retries_empty_final_response_with_summary_prompt():
 @pytest.mark.asyncio
 async def test_runner_uses_specific_message_after_empty_finalization_retry():
     """After silent retries + finalization all return empty, stop_reason is empty_final_response."""
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
     from nanobot.utils.runtime import EMPTY_FINAL_RESPONSE_MESSAGE
 
     provider = MagicMock(spec=LLMProvider)
@@ -295,7 +297,7 @@ async def test_runner_empty_response_does_not_break_tool_chain():
     Sequence: tool_call -> empty -> tool_call -> final text.
     The runner should recover via silent retry and complete normally.
     """
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     provider = MagicMock(spec=LLMProvider)
     call_count = 0
@@ -352,7 +354,7 @@ async def test_runner_empty_response_does_not_break_tool_chain():
 async def test_runner_accumulates_usage_and_preserves_cached_tokens():
     """Runner should accumulate prompt/completion tokens across iterations
     and preserve cached_tokens from provider responses."""
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     provider = MagicMock(spec=LLMProvider)
     call_count = {"n": 0}
@@ -399,7 +401,7 @@ async def test_runner_binds_on_retry_wait_to_retry_callback_not_progress():
     internal retry diagnostics like "Model request failed, retry in 1s"
     to leak to end-user channels as normal progress updates.
     """
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     captured: dict = {}
 
@@ -441,7 +443,7 @@ async def test_runner_binds_on_retry_wait_to_retry_callback_not_progress():
 @pytest.mark.asyncio
 async def test_runner_passes_temperature_to_provider():
     """temperature from AgentRunSpec should reach provider.chat_with_retry."""
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     captured: dict = {}
 
@@ -470,7 +472,7 @@ async def test_runner_passes_temperature_to_provider():
 @pytest.mark.asyncio
 async def test_runner_passes_max_tokens_to_provider():
     """max_tokens from AgentRunSpec should reach provider.chat_with_retry."""
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     captured: dict = {}
 
@@ -499,7 +501,7 @@ async def test_runner_passes_max_tokens_to_provider():
 @pytest.mark.asyncio
 async def test_runner_passes_reasoning_effort_to_provider():
     """reasoning_effort from AgentRunSpec should reach provider.chat_with_retry."""
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+    from nanobot.agent.runner import AgentRunner, AgentRunSpec
 
     captured: dict = {}