refactor(tools): use structured tool error results

tests/agent/test_runner_errors.py

@@ -135,6 +135,46 @@ async def test_runner_tool_error_sets_final_content():
     assert result.stop_reason == "tool_error"
 
 
+@pytest.mark.asyncio
+async def test_runner_preserves_successful_exec_output_that_starts_with_error():
+    from nanobot.agent.runner import AgentRunSpec, AgentRunner
+
+    provider = MagicMock(spec=LLMProvider)
+
+    async def chat_with_retry(*, messages, **kwargs):
+        if not any(msg.get("role") == "tool" for msg in messages):
+            return LLMResponse(
+                content="working",
+                tool_calls=[
+                    ToolCallRequest(id="call_1", name="exec", arguments={"command": "report"})
+                ],
+                usage={},
+            )
+        return LLMResponse(content="done", usage={})
+
+    provider.chat_with_retry = chat_with_retry
+    output = "Error: generated report successfully\n\nExit code: 0"
+    tools = MagicMock()
+    tools.get_definitions.return_value = []
+    tools.execute = AsyncMock(return_value=output)
+
+    runner = AgentRunner(provider)
+    result = await runner.run(AgentRunSpec(
+        initial_messages=[{"role": "user", "content": "run report"}],
+        tools=tools,
+        model="test-model",
+        max_iterations=2,
+        max_tool_result_chars=_MAX_TOOL_RESULT_CHARS,
+        fail_on_tool_error=True,
+    ))
+
+    assert result.final_content == "done"
+    assert result.stop_reason == "completed"
+    assert result.tool_events == [
+        {"name": "exec", "status": "ok", "detail": "Error: generated report successfully  Exit code: 0"}
+    ]
+
+
 @pytest.mark.asyncio
 async def test_runner_tool_error_preserves_tool_results_in_messages():
     """When a tool raises a fatal error, its results must still be appended

tests/agent/test_runner_safety.py

@@ -6,6 +6,8 @@ from unittest.mock import AsyncMock, MagicMock
 
 import pytest
 
+from nanobot.agent.runner import AgentRunner, AgentRunSpec
+from nanobot.agent.tools import ToolResult
 from nanobot.config.schema import AgentDefaults
 from nanobot.providers.base import LLMResponse, ToolCallRequest
 
@@ -20,8 +22,6 @@ async def test_runner_does_not_abort_on_workspace_violation_anymore():
     we now hand the error back to the LLM as a recoverable tool result and
     rely on ``repeated_workspace_violation_error`` to throttle bypass loops.
     """
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
-
     provider = MagicMock()
     provider.chat_with_retry = AsyncMock(side_effect=[
         LLMResponse(
@@ -64,8 +64,6 @@ async def test_runner_does_not_abort_on_workspace_violation_anymore():
 
 def test_is_ssrf_violation_recognizes_private_url_blocks():
     """SSRF rejections are classified separately from workspace boundaries."""
-    from nanobot.agent.runner import AgentRunner
-
     ssrf_msg = "Error: Command blocked by safety guard (internal/private URL detected)"
     assert AgentRunner._is_ssrf_violation(ssrf_msg) is True
     assert AgentRunner._is_ssrf_violation(
@@ -88,8 +86,6 @@ def test_is_ssrf_violation_recognizes_private_url_blocks():
 @pytest.mark.asyncio
 async def test_runner_returns_non_retryable_hint_on_ssrf_violation():
     """SSRF stays blocked, but the runtime gives the LLM a final chance to recover."""
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
-
     provider = MagicMock()
     provider.chat_with_retry = AsyncMock(side_effect=[
         LLMResponse(
@@ -107,7 +103,7 @@ async def test_runner_returns_non_retryable_hint_on_ssrf_violation():
     ])
     tools = MagicMock()
     tools.get_definitions.return_value = []
-    tools.execute = AsyncMock(return_value=(
+    tools.execute = AsyncMock(return_value=ToolResult.error(
         "Error: Command blocked by safety guard (internal/private URL detected)"
     ))
 
@@ -141,8 +137,6 @@ async def test_runner_lets_llm_recover_from_shell_guard_path_outside():
     turn (silent hang on Telegram per #3605); now the LLM gets the soft
     error back and can finalize on the next iteration.
     """
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
-
     provider = MagicMock()
     captured_second_call: list[dict] = []
 
@@ -163,7 +157,9 @@ async def test_runner_lets_llm_recover_from_shell_guard_path_outside():
     tools = MagicMock()
     tools.get_definitions.return_value = []
     tools.execute = AsyncMock(
-        return_value="Error: Command blocked by safety guard (path outside working dir)"
+        return_value=ToolResult.error(
+            "Error: Command blocked by safety guard (path outside working dir)"
+        )
     )
 
     runner = AgentRunner(provider)
@@ -195,8 +191,6 @@ async def test_runner_throttles_repeated_workspace_bypass_attempts():
     the runner replaces the tool result with a hard "stop trying" message
     so the model finally gives up and surfaces the boundary to the user.
     """
-    from nanobot.agent.runner import AgentRunSpec, AgentRunner
-
     bypass_attempts = [
         ToolCallRequest(
             id=f"a{i}", name="exec",
@@ -215,7 +209,9 @@ async def test_runner_throttles_repeated_workspace_bypass_attempts():
     tools = MagicMock()
     tools.get_definitions.return_value = []
     tools.execute = AsyncMock(
-        return_value="Error: Command blocked by safety guard (path outside working dir)"
+        return_value=ToolResult.error(
+            "Error: Command blocked by safety guard (path outside working dir)"
+        )
     )
 
     runner = AgentRunner(provider)