From 19a1d03fc881afc4d78e2f45990bf52e79896e1a Mon Sep 17 00:00:00 2001
From: hetaoBackend <hetao7@pku.edu.cn>
Date: Sun, 8 Feb 2026 21:09:18 +0800
Subject: [PATCH] fix: fix sub agent timeout

---
 backend/src/subagents/config.py         |  4 ++--
 backend/src/subagents/executor.py       |  3 ++-
 backend/src/tools/builtins/task_tool.py | 16 +++++++++++-----
 3 files changed, 15 insertions(+), 8 deletions(-)

diff --git a/backend/src/subagents/config.py b/backend/src/subagents/config.py
index cf0394c..8554e7d 100644
--- a/backend/src/subagents/config.py
+++ b/backend/src/subagents/config.py
@@ -15,7 +15,7 @@ class SubagentConfig:
         disallowed_tools: Optional list of tool names to deny.
         model: Model to use - 'inherit' uses parent's model.
         max_turns: Maximum number of agent turns before stopping.
-        timeout_seconds: Maximum execution time in seconds (default: 300 = 5 minutes).
+        timeout_seconds: Maximum execution time in seconds (default: 900 = 15 minutes).
     """
 
     name: str
@@ -25,4 +25,4 @@ class SubagentConfig:
     disallowed_tools: list[str] | None = field(default_factory=lambda: ["task"])
     model: str = "inherit"
     max_turns: int = 50
-    timeout_seconds: int = 300
+    timeout_seconds: int = 900
diff --git a/backend/src/subagents/executor.py b/backend/src/subagents/executor.py
index 65e5d81..b58aa9a 100644
--- a/backend/src/subagents/executor.py
+++ b/backend/src/subagents/executor.py
@@ -29,6 +29,7 @@ class SubagentStatus(Enum):
     RUNNING = "running"
     COMPLETED = "completed"
     FAILED = "failed"
+    TIMED_OUT = "timed_out"
 
 
 @dataclass
@@ -332,7 +333,7 @@ class SubagentExecutor:
                         f"[trace={self.trace_id}] Subagent {self.config.name} execution timed out after {self.config.timeout_seconds}s"
                     )
                     with _background_tasks_lock:
-                        _background_tasks[task_id].status = SubagentStatus.FAILED
+                        _background_tasks[task_id].status = SubagentStatus.TIMED_OUT
                         _background_tasks[task_id].error = f"Execution timed out after {self.config.timeout_seconds} seconds"
                         _background_tasks[task_id].completed_at = datetime.now()
                     # Cancel the future (best effort - may not stop the actual execution)
diff --git a/backend/src/tools/builtins/task_tool.py b/backend/src/tools/builtins/task_tool.py
index 44a86a8..36bf2ae 100644
--- a/backend/src/tools/builtins/task_tool.py
+++ b/backend/src/tools/builtins/task_tool.py
@@ -131,7 +131,7 @@ def task_tool(
             logger.info(f"[trace={trace_id}] Task {task_id} status: {result.status.value}")
             last_status = result.status
 
-        # Check if task completed or failed
+        # Check if task completed, failed, or timed out
         if result.status == SubagentStatus.COMPLETED:
             writer({"type": "task_completed", "task_id": task_id, "result": result.result})
             logger.info(f"[trace={trace_id}] Task {task_id} completed after {poll_count} polls")
@@ -140,14 +140,20 @@ def task_tool(
             writer({"type": "task_failed", "task_id": task_id, "error": result.error})
             logger.error(f"[trace={trace_id}] Task {task_id} failed: {result.error}")
             return f"Task failed. Error: {result.error}"
+        elif result.status == SubagentStatus.TIMED_OUT:
+            writer({"type": "task_timed_out", "task_id": task_id, "error": result.error})
+            logger.warning(f"[trace={trace_id}] Task {task_id} timed out: {result.error}")
+            return f"Task timed out. Error: {result.error}"
 
         # Still running, wait before next poll
         writer({"type": "task_running", "task_id": task_id, "poll_count": poll_count})
         time.sleep(5)  # Poll every 5 seconds
         poll_count += 1
 
-        # Optional: Add timeout protection (e.g., max 5 minutes)
-        if poll_count > 60:  # 60 * 5s = 5 minutes
-            logger.warning(f"[trace={trace_id}] Task {task_id} timed out after {poll_count} polls")
+        # Polling timeout as a safety net (in case thread pool timeout doesn't work)
+        # Set to 16 minutes (longer than the default 15-minute thread pool timeout)
+        # This catches edge cases where the background task gets stuck
+        if poll_count > 192:  # 192 * 5s = 16 minutes
+            logger.error(f"[trace={trace_id}] Task {task_id} polling timed out after {poll_count} polls (should have been caught by thread pool timeout)")
             writer({"type": "task_timed_out", "task_id": task_id})
-            return f"Task timed out after 5 minutes. Status: {result.status.value}"
+            return f"Task polling timed out after 16 minutes. This may indicate the background task is stuck. Status: {result.status.value}"