fix(intelligence): Heartbeat 统一健康系统 — 6处断链修复 + 健康面板 + SaaS自动恢复

Rust 后端 (heartbeat.rs):
- 告警实时推送: OnceLock<AppHandle> + Tauri emit heartbeat:alert
- 动态间隔: tokio::select! + Notify 替代不可变 interval
- Config 持久化: update_config 写入 VikingStorage
- heartbeat_init 从 VikingStorage 恢复 config
- 移除 dead code (subscribe, HeartbeatCheckFn)
- Memory stats fallback 分层处理

新增 health_snapshot.rs:
- HealthSnapshot Tauri 命令 — 按需查询引擎/记忆状态
- 注册到 lib.rs invoke_handler

前端修复:
- HeartbeatConfig handleSave 同步到 Rust 后端
- App.tsx 读 localStorage 持久化配置 + heartbeat:alert 监听 + toast
- saasStore 降级后指数退避探测恢复 + saas-recovered 事件
- 新增 HealthPanel.tsx 只读健康面板 (4卡片 + 告警列表)
- SettingsLayout 添加 health 导航入口

清理:
- 删除 intelligence-client/ 目录版 (9文件 -1640行, 单文件版是活跃代码)

desktop/src/store/saasStore.ts

@@ -84,6 +84,7 @@ export interface SaaSStateSlice {
   _consecutiveFailures: number;
   _heartbeatTimer?: ReturnType<typeof setInterval>;
   _healthCheckTimer?: ReturnType<typeof setInterval>;
+  _recoveryProbeTimer?: ReturnType<typeof setInterval>;
 
   // === Billing State ===
   plans: BillingPlan[];
@@ -141,6 +142,67 @@ function resolveInitialMode(sessionMeta: { saasUrl: string; account: SaaSAccount
   return sessionMeta ? 'saas' : 'tauri';
 }
 
+// === SaaS Recovery Probe ===
+// When SaaS degrades to local mode, periodically probes SaaS reachability
+// with exponential backoff (2min → 3min → 4.5min → 6.75min → 10min cap).
+// On recovery, switches back to SaaS mode and notifies user via toast.
+
+let _recoveryProbeInterval: ReturnType<typeof setInterval> | null = null;
+let _recoveryBackoffMs = 2 * 60 * 1000; // Start at 2 minutes
+const RECOVERY_BACKOFF_CAP_MS = 10 * 60 * 1000; // Max 10 minutes
+const RECOVERY_BACKOFF_MULTIPLIER = 1.5;
+
+function startRecoveryProbe() {
+  if (_recoveryProbeInterval) return; // Already probing
+
+  _recoveryBackoffMs = 2 * 60 * 1000; // Reset backoff
+  log.info('[SaaS Recovery] Starting recovery probe...');
+
+  const probe = async () => {
+    try {
+      await saasClient.deviceHeartbeat(DEVICE_ID);
+      // SaaS is reachable again — recover
+      log.info('[SaaS Recovery] SaaS reachable — switching back to SaaS mode');
+      useSaaSStore.setState({
+        saasReachable: true,
+        connectionMode: 'saas',
+        _consecutiveFailures: 0,
+      } as unknown as Partial<SaaSStore>);
+      saveConnectionMode('saas');
+
+      // Notify user via custom event (App.tsx listens)
+      if (typeof window !== 'undefined') {
+        window.dispatchEvent(new CustomEvent('saas-recovered'));
+      }
+
+      // Stop probing
+      stopRecoveryProbe();
+    } catch {
+      // Still unreachable — increase backoff
+      _recoveryBackoffMs = Math.min(
+        _recoveryBackoffMs * RECOVERY_BACKOFF_MULTIPLIER,
+        RECOVERY_BACKOFF_CAP_MS
+      );
+      log.debug(`[SaaS Recovery] Still unreachable, next probe in ${Math.round(_recoveryBackoffMs / 1000)}s`);
+
+      // Reschedule with new backoff
+      if (_recoveryProbeInterval) {
+        clearInterval(_recoveryProbeInterval);
+      }
+      _recoveryProbeInterval = setInterval(probe, _recoveryBackoffMs);
+    }
+  };
+
+  _recoveryProbeInterval = setInterval(probe, _recoveryBackoffMs);
+}
+
+function stopRecoveryProbe() {
+  if (_recoveryProbeInterval) {
+    clearInterval(_recoveryProbeInterval);
+    _recoveryProbeInterval = null;
+  }
+}
+
 // === Store Implementation ===
 
 export const useSaaSStore = create<SaaSStore>((set, get) => {
@@ -698,6 +760,8 @@ export const useSaaSStore = create<SaaSStore>((set, get) => {
                   connectionMode: 'tauri',
                 } as unknown as Partial<SaaSStore>);
                 saveConnectionMode('tauri');
+                // Start recovery probe with exponential backoff
+                startRecoveryProbe();
               }
             }
           }, 5 * 60 * 1000);