feat: support uploaded images in debug voice preview

Allow paste/drag temporary image assets so vision turns work without a live camera frame.

Co-authored-by: Cursor <cursoragent@cursor.com>

backend/tests/test_user_input.py

@@ -32,7 +32,7 @@ class UserInputParserTests(unittest.TestCase):
         self.assertIsNotNone(value)
         self.assertEqual(value.text, "帮我看看")
         self.assertTrue(value.has_camera_frame)
-        self.assertEqual(value.transcript_text, "帮我看看")
+        self.assertEqual(value.transcript_text, "")
 
     def test_rejects_legacy_and_unsupported_image_sources(self):
         self.assertIsNone(parse_user_input({"type": "user-text", "text": "旧协议"}))
@@ -46,14 +46,40 @@ class UserInputParserTests(unittest.TestCase):
                         {
                             "type": "input_image",
                             "source": {
-                                "type": "uploaded_asset",
-                                "asset_id": "asset_1",
+                                "type": "remote_url",
+                                "url": "https://example.com/image.jpg",
                             },
                         }
                     ],
                 }
             )
 
+    def test_parses_uploaded_image_asset(self):
+        value = parse_user_input(
+            {
+                "type": "user-input",
+                "schema_version": 1,
+                "input_id": "input_upload",
+                "parts": [
+                    {"type": "input_text", "text": "这是什么？"},
+                    {
+                        "type": "input_image",
+                        "source": {
+                            "type": "uploaded_asset",
+                            "asset_token": "signed-token",
+                        },
+                    },
+                ],
+            }
+        )
+
+        self.assertIsNotNone(value)
+        self.assertTrue(value.has_image)
+        self.assertFalse(value.has_camera_frame)
+        self.assertEqual(value.image_asset_token, "signed-token")
+        self.assertEqual(value.prompt_text, "这是什么？")
+        self.assertEqual(value.transcript_text, "")
+
     def test_image_only_input_has_no_synthetic_chat_text(self):
         value = parse_user_input(
             {