feat: add clear error message when llama-server not running

src/models/local/llamacpp.test.ts

@@ -81,4 +81,16 @@ describe('LlamaCppClient', () => {
       usage: { inputTokens: 5, outputTokens: 2 },
     });
   });
+
+  it('throws clear error when server not running', async () => {
+    mockFetch.mockRejectedValue(new TypeError('fetch failed'));
+
+    const client = new LlamaCppClient({
+      endpoint: 'http://localhost:8080',
+    });
+
+    await expect(client.chat({
+      messages: [{ role: 'user', content: 'Hello' }],
+    })).rejects.toThrow('llama-server not running at http://localhost:8080');
+  });
 });