fix(long-task): honor final signal and file tracking

Co-authored-by: Cursor <cursoragent@cursor.com>

nanobot/agent/tools/long_task.py

@@ -4,7 +4,7 @@ from __future__ import annotations
 
 import time
 from dataclasses import dataclass, field
-from typing import Any, TYPE_CHECKING
+from typing import TYPE_CHECKING, Any
 
 from loguru import logger
 
@@ -30,6 +30,7 @@ if TYPE_CHECKING:
 class HandoffState:
     """Structured progress state passed between long-task steps."""
 
+    signal_type: str = ""
     message: str = ""
     files_created: list[str] = field(default_factory=list)
     files_modified: list[str] = field(default_factory=list)
@@ -39,6 +40,7 @@ class HandoffState:
     def is_empty(self) -> bool:
         return not any(
             [
+                self.signal_type,
                 self.message,
                 self.files_created,
                 self.files_modified,
@@ -105,6 +107,7 @@ class HandoffTool(Tool):
         verification: str = "",
         **kwargs: Any,
     ) -> str:
+        self._store.signal_type = "handoff"
         self._store.message = message
         self._store.files_created = list(files_created or [])
         self._store.files_modified = list(files_modified or [])
@@ -139,6 +142,7 @@ class CompleteTool(Tool):
         )
 
     async def execute(self, summary: str, **kwargs: Any) -> str:
+        self._store.signal_type = "complete"
         self._store.message = summary
         return "Task marked as complete. Awaiting validation."
 
@@ -242,9 +246,19 @@ def _extract_handoff_from_messages(messages: list[dict[str, Any]]) -> str:
     return ""
 
 
-_FILE_EVENT_PREFIXES = ("Wrote ", "Edited ")
+def _extract_write_path(detail: str) -> str | None:
-# NOTE: path extraction depends on write_file/edit_file detail format.
+    prefix = "Successfully wrote "
-# If those tools change their output format, this mapping must be updated.
+    marker = " to "
+    if not detail.startswith(prefix) or marker not in detail:
+        return None
+    return detail.rsplit(marker, 1)[1].strip()
+
+
+def _extract_edit_path(detail: str) -> str | None:
+    for prefix in ("Successfully created ", "Successfully edited "):
+        if detail.startswith(prefix):
+            return detail.removeprefix(prefix).strip()
+    return None
 
 
 def _extract_file_changes(
@@ -259,13 +273,19 @@ def _extract_file_changes(
         detail = event.get("detail", "")
         if status != "ok":
             continue
-        if name in ("write_file", "edit_file"):
+        if name == "write_file":
-            if detail.startswith(_FILE_EVENT_PREFIXES):
+            path = _extract_write_path(detail)
-                path = detail.split(" ", 1)[1].split(":")[0].strip()
+            if path:
-                if name == "write_file":
+                created.append(path)
-                    created.append(path)
+            else:
-                else:
+                logger.debug(
-                    modified.append(path)
+                    "long_task: skipping file event with unexpected detail: {}",
+                    detail[:80],
+                )
+        elif name == "edit_file":
+            path = _extract_edit_path(detail)
+            if path:
+                modified.append(path)
             else:
                 logger.debug(
                     "long_task: skipping file event with unexpected detail: {}",
@@ -502,7 +522,7 @@ class LongTaskTool(Tool):
 
             # Determine signal from tool events
             sig_type = "none"
-            for event in tool_events:
+            for event in reversed(tool_events):
                 ev_name = event.get("name", "")
                 if ev_name == "complete":
                     sig_type = "complete"
@@ -512,11 +532,9 @@ class LongTaskTool(Tool):
                     break
 
             # Fallback: if no explicit signal but CompleteTool/HandoffTool was
-            # called without arguments (message empty), use final_content
+            # called but the runner did not expose tool events, trust the store.
-            if sig_type == "none" and signal_store.message:
+            if sig_type == "none" and signal_store.signal_type:
-                # Tool was called but we couldn't detect from events;
+                sig_type = signal_store.signal_type
-                # use the store content as handoff
-                sig_type = "handoff"
             elif sig_type == "none":
                 signal_store.message = _extract_handoff_from_messages(
                     getattr(result, "messages", []) or []

tests/agent/tools/test_long_task.py

@@ -1,22 +1,21 @@
 """Tests for Long Task Tool: HandoffTool, CompleteTool, LongTaskTool."""
 
-import pytest
 from types import SimpleNamespace
 from unittest.mock import AsyncMock, MagicMock
 
+import pytest
+
 from nanobot.agent.tools.long_task import (
+    CompleteTool,
     HandoffState,
     HandoffTool,
-    CompleteTool,
     LongTaskTool,
-    LongTaskEvent,
     _build_system_prompt,
     _build_user_message,
     _extract_file_changes,
     _extract_handoff_from_messages,
 )
 
-
 # ---------------------------------------------------------------------------
 # Signal tool tests
 # ---------------------------------------------------------------------------
@@ -34,6 +33,7 @@ async def test_handoff_tool_stores_structured_signal():
         verification="Tests passed",
     )
     assert result == "Progress recorded. The next step will continue from here."
+    assert store.signal_type == "handoff"
     assert store.message == "Processed items 1-8. Results in out.md."
     assert store.files_created == ["out.md", "report.md"]
     assert store.files_modified == ["main.py"]
@@ -58,6 +58,7 @@ async def test_complete_tool_stores_signal():
     tool = CompleteTool(store)
     result = await tool.execute(summary="All 100 items processed. Summary in report.md")
     assert result == "Task marked as complete. Awaiting validation."
+    assert store.signal_type == "complete"
     assert store.message == "All 100 items processed. Summary in report.md"
 
 
@@ -70,6 +71,7 @@ async def test_signal_tools_overwrite_on_multiple_calls():
     await handoff.execute(message="first progress")
     assert store.message == "first progress"
     await complete.execute(summary="done early")
+    assert store.signal_type == "complete"
     assert store.message == "done early"
 
 
@@ -191,6 +193,44 @@ async def test_long_task_completes_after_multiple_handoffs():
     assert call_count == 4  # 3 main steps + validation
 
 
+@pytest.mark.asyncio
+async def test_long_task_uses_last_signal_when_multiple_signals_called():
+    """If a step calls handoff() then complete(), complete() should win."""
+    mgr = _make_manager_stub()
+    call_count = 0
+
+    async def fake_run_step(*, system_prompt, user_message, extra_tools, max_iterations=None):
+        nonlocal call_count
+        call_count += 1
+        if call_count == 1:
+            for t in extra_tools:
+                if t.name == "handoff":
+                    await t.execute(message="Partial progress.")
+                elif t.name == "complete":
+                    await t.execute(summary="Actually complete.")
+            return _step_result(
+                tools_used=["handoff", "complete"],
+                tool_events=[
+                    {"name": "handoff", "status": "ok", "detail": ""},
+                    {"name": "complete", "status": "ok", "detail": ""},
+                ],
+            )
+
+        for t in extra_tools:
+            if t.name == "complete":
+                await t.execute(summary="Validated")
+        return _step_result(
+            tools_used=["complete"],
+            tool_events=[{"name": "complete", "status": "ok", "detail": ""}],
+        )
+
+    mgr.run_step.side_effect = fake_run_step
+    tool = LongTaskTool(manager=mgr)
+    result = await tool.execute(goal="Do something.", max_steps=1)
+    assert result == "Actually complete."
+    assert call_count == 2  # main step + validation
+
+
 @pytest.mark.asyncio
 async def test_long_task_validation_falls_back_to_handoff():
     """Subagent claims complete but validation fails — task continues."""
@@ -566,14 +606,27 @@ def test_extract_handoff_from_empty_messages():
 
 def test_extract_file_changes_from_tool_events():
     events = [
-        {"name": "write_file", "status": "ok", "detail": "Wrote /workspace/a.py: done"},
+        {
-        {"name": "edit_file", "status": "ok", "detail": "Edited /workspace/b.py: patched"},
+            "name": "write_file",
+            "status": "ok",
+            "detail": "Successfully wrote 12 characters to /workspace/a.py",
+        },
+        {
+            "name": "edit_file",
+            "status": "ok",
+            "detail": "Successfully edited /workspace/b.py",
+        },
+        {
+            "name": "edit_file",
+            "status": "ok",
+            "detail": "Successfully created /workspace/c.py",
+        },
         {"name": "read_file", "status": "ok", "detail": "Read /workspace/c.py"},
         {"name": "write_file", "status": "error", "detail": "Failed"},
     ]
     created, modified = _extract_file_changes(events)
     assert created == ["/workspace/a.py"]
-    assert modified == ["/workspace/b.py"]
+    assert modified == ["/workspace/b.py", "/workspace/c.py"]
 
 
 def test_extract_file_changes_empty():
@@ -736,7 +789,11 @@ async def test_explicit_file_changes_override_auto_detected():
             tool_events=[
                 {"name": "handoff", "status": "ok", "detail": ""},
                 # Auto-detection would pick this up as "auto.py"
-                {"name": "write_file", "status": "ok", "detail": "Wrote auto.py: content"},
+                {
+                    "name": "write_file",
+                    "status": "ok",
+                    "detail": "Successfully wrote 7 characters to auto.py",
+                },
             ],
         )