Phase 4: backups w/ volumes, notifications, settings & users, audit page (0.4.0)

- Backup/restore: per-stack tar.gz incl. named-volume snapshots (helper
  container), upload restore with rename/overwrite/conflict detection.
- Notifications: ntfy/Discord/Slack/Gotify/generic webhooks, per-event
  subscriptions; wired into the update checker and stack lifecycle.
- Settings page: update-check interval, webhook CRUD + test, user management
  (with last-admin safeguards).
- Audit log page (searchable, paginated).
- Mobile-responsive sidebar/layout.

Multi-host agents and remote backup destinations (SFTP/S3) deferred.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

backend/services/update_service.py

@@ -16,6 +16,8 @@ import httpx
 
 from config import settings
 from docker_client import DockerError, get_client, safe_call
+from models.setting import EVENT_UPDATE_AVAILABLE
+from services import notify_service, settings_service
 
 logger = logging.getLogger("stackpilot.update")
 
@@ -45,6 +47,10 @@ class UpdateStatus:
 # image ref -> UpdateStatus
 _CACHE: dict[str, UpdateStatus] = {}
 
+# images we've already sent an "update available" notification for, so the
+# background loop doesn't re-notify on every cycle.
+_NOTIFIED: set[str] = set()
+
 
 # --------------------------------------------------------------------------- #
 # Image reference parsing
@@ -164,6 +170,18 @@ async def check_image(image: str) -> UpdateStatus:
         error=error,
     )
     _CACHE[image] = status
+    if update_available and image not in _NOTIFIED:
+        _NOTIFIED.add(image)
+        try:
+            await notify_service.notify(
+                EVENT_UPDATE_AVAILABLE,
+                "Image update available",
+                f"A newer image is available for {image}.",
+            )
+        except Exception as exc:  # noqa: BLE001 - notifications are best-effort
+            logger.debug("update notify failed for %s: %s", image, exc)
+    elif not update_available:
+        _NOTIFIED.discard(image)
     return status
 
 
@@ -192,7 +210,6 @@ def get_cache() -> dict[str, dict]:
 
 
 async def background_loop():
-    interval = max(settings.UPDATE_CHECK_INTERVAL_MINUTES, 5) * 60
     # initial delay so startup isn't blocked
     await asyncio.sleep(30)
     while True:
@@ -201,4 +218,6 @@ async def background_loop():
             logger.info("Image update check complete (%d images)", len(_CACHE))
         except Exception as exc:  # noqa: BLE001
             logger.warning("Image update check failed: %s", exc)
+        # Re-read the interval each cycle so Settings changes take effect.
+        interval = max(settings_service.get_update_interval(), 5) * 60
         await asyncio.sleep(interval)