feat: consolidate checkpoint metadata handling

- Extract checkpoint entry from multiple metadata locations using helper method
- Sanitize checkpoint metadata by removing transient/local-only fields
- Remove checkpoint duplication from generation parameters to store only at top level
- Update frontend to properly populate checkpoint metadata during import
- Add tests for new checkpoint handling functionality

This ensures consistent checkpoint metadata structure and prevents data duplication across different storage locations.

py/services/recipes/persistence_service.py

@@ -79,7 +79,7 @@ class RecipePersistenceService:
 
         current_time = time.time()
         loras_data = [self._normalise_lora_entry(lora) for lora in (metadata.get("loras") or [])]
-        checkpoint_entry = metadata.get("checkpoint")
+        checkpoint_entry = self._sanitize_checkpoint_entry(self._extract_checkpoint_entry(metadata))
 
         gen_params = metadata.get("gen_params") or {}
         if not gen_params and "raw_metadata" in metadata:
@@ -87,7 +87,6 @@ class RecipePersistenceService:
             gen_params = {
                 "prompt": raw_metadata.get("prompt", ""),
                 "negative_prompt": raw_metadata.get("negative_prompt", ""),
-                "checkpoint": raw_metadata.get("checkpoint", {}),
                 "steps": raw_metadata.get("steps", ""),
                 "sampler": raw_metadata.get("sampler", ""),
                 "cfg_scale": raw_metadata.get("cfg_scale", ""),
@@ -95,8 +94,9 @@ class RecipePersistenceService:
                 "size": raw_metadata.get("size", ""),
                 "clip_skip": raw_metadata.get("clip_skip", ""),
             }
-        if checkpoint_entry and "checkpoint" not in gen_params:
-            gen_params["checkpoint"] = checkpoint_entry
+
+        # Drop checkpoint duplication from generation parameters to store it only at top level
+        gen_params.pop("checkpoint", None)
 
         fingerprint = calculate_recipe_fingerprint(loras_data)
         recipe_data: Dict[str, Any] = {
@@ -335,7 +335,7 @@ class RecipePersistenceService:
             "created_date": time.time(),
             "base_model": most_common_base_model,
             "loras": loras_data,
-            "checkpoint": metadata.get("checkpoint", ""),
+            "checkpoint": self._sanitize_checkpoint_entry(metadata.get("checkpoint", "")),
             "gen_params": {
                 key: value
                 for key, value in metadata.items()
@@ -364,6 +364,30 @@ class RecipePersistenceService:
 
     # Helper methods ---------------------------------------------------
 
+    def _extract_checkpoint_entry(self, metadata: dict[str, Any]) -> Optional[dict[str, Any]]:
+        """Pull a checkpoint entry from various metadata locations."""
+
+        checkpoint_entry = metadata.get("checkpoint") or metadata.get("model")
+        if not checkpoint_entry:
+            gen_params = metadata.get("gen_params") or {}
+            checkpoint_entry = gen_params.get("checkpoint")
+
+        return checkpoint_entry if isinstance(checkpoint_entry, dict) else None
+
+    def _sanitize_checkpoint_entry(self, checkpoint_entry: Optional[dict[str, Any]]) -> Optional[dict[str, Any]]:
+        """Remove transient/local-only fields from checkpoint metadata."""
+
+        if not checkpoint_entry:
+            return None
+
+        if not isinstance(checkpoint_entry, dict):
+            return checkpoint_entry
+
+        pruned = dict(checkpoint_entry)
+        for key in ("existsLocally", "localPath", "thumbnailUrl", "size", "downloadUrl"):
+            pruned.pop(key, None)
+        return pruned
+
     def _resolve_image_bytes(self, image_bytes: bytes | None, image_base64: str | None) -> bytes:
         if image_bytes is not None:
             return image_bytes

static/js/managers/import/DownloadManager.js

@@ -56,6 +56,15 @@ export class DownloadManager {
                     gen_params: this.importManager.recipeData.gen_params || {},
                     raw_metadata: this.importManager.recipeData.raw_metadata || {}
                 };
+
+                const checkpointMetadata =
+                    this.importManager.recipeData.checkpoint ||
+                    this.importManager.recipeData.model ||
+                    (this.importManager.recipeData.gen_params || {}).checkpoint;
+
+                if (checkpointMetadata && typeof checkpointMetadata === 'object') {
+                    completeMetadata.checkpoint = checkpointMetadata;
+                }
                 
                 // Add source_path to metadata to track where the recipe was imported from
                 if (this.importManager.importMode === 'url') {

tests/services/test_recipe_services.py

@@ -203,7 +203,117 @@ async def test_save_recipe_persists_checkpoint_metadata(tmp_path):
 
     stored = json.loads(Path(result.payload["json_path"]).read_text())
     assert stored["checkpoint"] == checkpoint_meta
-    assert stored["gen_params"]["checkpoint"] == checkpoint_meta
+    assert "checkpoint" not in stored["gen_params"]
+
+
+@pytest.mark.asyncio
+async def test_save_recipe_promotes_checkpoint_from_gen_params(tmp_path):
+    exif_utils = DummyExifUtils()
+
+    class DummyScanner:
+        def __init__(self, root):
+            self.recipes_dir = str(root)
+
+        async def find_recipes_by_fingerprint(self, fingerprint):
+            return []
+
+        async def add_recipe(self, recipe_data):
+            return None
+
+    scanner = DummyScanner(tmp_path)
+    service = RecipePersistenceService(
+        exif_utils=exif_utils,
+        card_preview_width=512,
+        logger=logging.getLogger("test"),
+    )
+
+    checkpoint_meta = {
+        "type": "checkpoint",
+        "modelId": 10,
+        "modelVersionId": 20,
+        "modelName": "Flux",
+        "modelVersionName": "Dev",
+    }
+
+    metadata = {
+        "base_model": "Flux",
+        "loras": [],
+        "gen_params": {
+            "checkpoint": checkpoint_meta,
+        },
+    }
+
+    result = await service.save_recipe(
+        recipe_scanner=scanner,
+        image_bytes=b"img",
+        image_base64=None,
+        name="Checkpointed",
+        tags=[],
+        metadata=metadata,
+    )
+
+    stored = json.loads(Path(result.payload["json_path"]).read_text())
+    assert stored["checkpoint"] == checkpoint_meta
+    assert "checkpoint" not in stored["gen_params"]
+
+
+@pytest.mark.asyncio
+async def test_save_recipe_strips_checkpoint_local_fields(tmp_path):
+    exif_utils = DummyExifUtils()
+
+    class DummyScanner:
+        def __init__(self, root):
+            self.recipes_dir = str(root)
+
+        async def find_recipes_by_fingerprint(self, fingerprint):
+            return []
+
+        async def add_recipe(self, recipe_data):
+            return None
+
+    scanner = DummyScanner(tmp_path)
+    service = RecipePersistenceService(
+        exif_utils=exif_utils,
+        card_preview_width=512,
+        logger=logging.getLogger("test"),
+    )
+
+    checkpoint_meta = {
+        "type": "checkpoint",
+        "modelId": 10,
+        "modelVersionId": 20,
+        "modelName": "Flux",
+        "modelVersionName": "Dev",
+        "existsLocally": False,
+        "localPath": "/tmp/foo",
+        "thumbnailUrl": "http://example.com",
+        "size": 123,
+        "downloadUrl": "http://example.com/dl",
+    }
+
+    metadata = {
+        "base_model": "Flux",
+        "loras": [],
+        "checkpoint": checkpoint_meta,
+    }
+
+    result = await service.save_recipe(
+        recipe_scanner=scanner,
+        image_bytes=b"img",
+        image_base64=None,
+        name="Checkpointed",
+        tags=[],
+        metadata=metadata,
+    )
+
+    stored = json.loads(Path(result.payload["json_path"]).read_text())
+    assert stored["checkpoint"] == {
+        "type": "checkpoint",
+        "modelId": 10,
+        "modelVersionId": 20,
+        "modelName": "Flux",
+        "modelVersionName": "Dev",
+    }
 
 
 @pytest.mark.asyncio