BUG: Daemon shutdown sequence has no goroutine coordination

[andrinoff]

Describe the bug

daemon.Run() starts 4 concurrent goroutines (idleEventLoop, handleSignals, backgroundSync, acceptLoop) but has no WaitGroup or coordination during shutdown. When Shutdown() fires:

• backgroundSync() may be mid-fetch when providers are closed
• idleEventLoop() may process a batch while connections are torn down
• Untracked goroutines spawned by fetchAndCache() race with cleanup
• No guarantee all goroutines finish before Run() returns

To reproduce

1. Run daemon with multiple accounts and active IDLE watchers
2. Send SIGTERM during a background sync
3. Race between sync goroutine and provider closure

Expected behavior

Add sync.WaitGroup to track all goroutines. Ensure Run() blocks until all goroutines complete. Cancel in-flight operations via context before closing providers.

[floatpanebot]

This has had no activity for 30 days. It will be closed in 14 days unless updated. Comment or remove the stale label to keep it open.

[floatpanebot]

Closing due to inactivity. Reopen if still relevant.