fix: harden L3 runtime continuity and tool execution

Align the L3 graph, agent service, and sync tool shims on one canonical continuity contract so clarification resumes and persisted snapshots behave consistently. Add targeted regressions and hardening notes covering system-message coalescing, async bridge usage, and continuity rehydration.

backend/app/agents/tools/schedule.py

@@ -2,8 +2,6 @@
 
 from __future__ import annotations
 
-import asyncio
-from concurrent.futures import ThreadPoolExecutor
 from datetime import date, datetime
 from zoneinfo import ZoneInfo
 
@@ -11,21 +9,16 @@ from langchain_core.tools import tool
 from sqlalchemy import select
 
 from app.agents.context import get_current_user
+from app.agents.tools.async_bridge import run_async
 from app.database import async_session
 from app.models.goal import Goal, GoalStatus
 from app.models.reminder import Reminder
 from app.models.task import Task, TaskPriority, TaskStatus
 from app.models.todo import DailyTodo, TodoSource
 
-_executor = ThreadPoolExecutor(max_workers=4)
-
 
 def _run_async(coro, timeout: int = 30):
-    try:
-        asyncio.get_running_loop()
-    except RuntimeError:
-        return asyncio.run(coro)
-    return _executor.submit(asyncio.run, coro).result(timeout=timeout)
+    return run_async(coro, timeout=timeout)
 
 
 def _parse_date(value: str | None) -> date: