feat(backend): auto-stop/start daemon when switching backends

- Add local_providers with ollama and llamacpp configurations
- /backend command now stops current daemon before starting new one
- Start backends as detached processes to avoid blocking TUI
- Wait 500ms for daemon to initialize before switching

src/frontends/tui/minimal.ts

@@ -194,7 +194,7 @@ export class MinimalTui {
         break;
 
       case 'backend':
-        this.handleBackendCommand(command.provider);
+        await this.handleBackendCommand(command.provider);
         break;
 
       case 'login':
@@ -277,7 +277,7 @@ export class MinimalTui {
     }
   }
 
-  private handleBackendCommand(provider?: string): void {
+  private async handleBackendCommand(provider?: string): Promise<void> {
     const router = this.config.modelRouter;
     if (!router) {
       console.log('Backend switching not available.\n');
@@ -300,6 +300,17 @@ export class MinimalTui {
       return;
     }
 
+    // Stop current daemon if running
+    const currentBackend = router.getLocalProviderName();
+    if (currentBackend && currentBackend !== provider) {
+      console.log(`${colors.gray}Stopping ${currentBackend}...${colors.reset}`);
+      await this.stopBackend(currentBackend);
+    }
+
+    // Start new daemon
+    console.log(`${colors.gray}Starting ${provider}...${colors.reset}`);
+    await this.startBackend(provider, providerConfig);
+
     const client = this.createLocalClient(providerConfig);
     if (!client) {
       console.log(`Failed to create client for '${provider}'.\n`);
@@ -307,7 +318,53 @@ export class MinimalTui {
     }
 
     router.setLocalClient(client, provider);
-    console.log(`Switched to backend: ${provider}\n`);
+    console.log(`${colors.gray}Switched to backend: ${provider}${colors.reset}\n`);
+  }
+
+  private async stopBackend(provider: string): Promise<void> {
+    try {
+      const { spawn } = await import('child_process');
+      let processName: string;
+      switch (provider) {
+        case 'ollama':
+          processName = 'ollama';
+          break;
+        case 'llamacpp':
+          processName = 'llama-server';
+          break;
+        default:
+          return;
+      }
+      await new Promise<void>((resolve) => {
+        spawn('pkill', [processName]).on('close', resolve);
+      });
+    } catch (error) {
+      // Ignore errors stopping backends
+    }
+  }
+
+  private async startBackend(provider: string, config: ModelConfig): Promise<void> {
+    try {
+      const { spawn } = await import('child_process');
+      const args: string[] = [];
+
+      switch (provider) {
+        case 'ollama':
+          spawn('ollama', ['serve'], { detached: true, stdio: 'ignore' }).unref();
+          break;
+        case 'llamacpp':
+          args.push('--model', config.model);
+          args.push('--port', new URL(config.endpoint ?? 'http://localhost:8080').port || '8080');
+          args.push('--host', '0.0.0.0');
+          spawn('llama-server', args, { detached: true, stdio: 'ignore' }).unref();
+          break;
+      }
+
+      // Wait briefly for the daemon to start
+      await new Promise(resolve => setTimeout(resolve, 500));
+    } catch (error) {
+      console.log(`${colors.gray}Warning: Failed to start ${provider}: ${error instanceof Error ? error.message : String(error)}${colors.reset}\n`);
+    }
   }
 
   private async handleLoginCommand(provider?: string): Promise<void> {