fix(test): sync civitai bulk test with nsfw param

CivitAI image API returns modelVersionIds at the root level of the
response (not inside meta), containing ALL model version IDs across
all resources (checkpoint + LoRAs). Two bugs prevented LoRAs from
being discovered:

1. _download_remote_media only extracted the first modelVersionId for
   enrichment, dropping the rest.
2. CivitAI API meta parsing only ran as an EXIF fallback, but most
   images have embedded EXIF metadata (prompt, steps, etc.), so the
   fallback was never triggered.
3. When civitai_meta_raw itself has a nested 'meta' key, unwrapping
   it stripped the injected modelVersionIds.

Also fixed gen_params merge: API gen_params now overlays EXIF at the
field level instead of full replacement, preserving EXIF-only fields
like detailed generation parameters.

py/routes/handlers/recipe_handlers.py

@@ -871,28 +871,47 @@ class RecipeManagementHandler:
                 "Failed to extract embedded metadata during import: %s", exc
             )
 
-        # Fallback: if EXIF extraction yielded nothing, parse Civitai API meta directly
+        # Parse CivitAI API meta to discover all resources from modelVersionIds
-        # (same approach as analyze_remote_image — downloaded Civitai images often
+        # (modelVersionIds is injected at root level by _download_remote_media).
-        # have no embedded EXIF but the API meta contains resources/hashes)
+        # Run unconditionally — EXIF parsing may succeed for gen_params but miss
-        if parsed_embedded is None and civitai_meta_raw:
+        # LoRAs since modelVersionIds is NOT embedded in the image EXIF.
+        civitai_parsed = None
+        if civitai_meta_raw:
             civitai_inner_meta = civitai_meta_raw
             if isinstance(civitai_meta_raw, dict) and "meta" in civitai_meta_raw:
                 civitai_inner_meta = civitai_meta_raw["meta"]
+                # modelVersionIds lives at outer meta level; propagate after unwrap
+                _mvids = civitai_meta_raw.get("modelVersionIds")
+                if _mvids and isinstance(civitai_inner_meta, dict):
+                    civitai_inner_meta["modelVersionIds"] = _mvids
             if isinstance(civitai_inner_meta, dict):
                 parser = self._analysis_service._recipe_parser_factory.create_parser(
                     civitai_inner_meta
                 )
                 if parser:
-                    parsed_embedded = await parser.parse_metadata(
+                    civitai_parsed = await parser.parse_metadata(
                         civitai_inner_meta, recipe_scanner=recipe_scanner
                     )
-                    if parsed_embedded and "gen_params" in parsed_embedded:
+                    if civitai_parsed and "gen_params" in civitai_parsed:
-                        embedded_gen_params = parsed_embedded["gen_params"]
+                        # Merge: API gen_params override EXIF at field level,
+                        # EXIF fills in fields the API doesn't have.
+                        embedded_gen_params = {
+                            **(embedded_gen_params or {}),
+                            **civitai_parsed["gen_params"],
+                        }
 
         if embedded_gen_params:
             metadata["gen_params"] = embedded_gen_params
 
-        if parsed_embedded:
+        # Merge LoRAs: prefer frontend resources, supplement with CivitAI modelVersionIds
+        if civitai_parsed:
+            civitai_loras = civitai_parsed.get("loras", [])
+            if civitai_loras and not metadata.get("loras"):
+                metadata["loras"] = civitai_loras
+            civitai_model = civitai_parsed.get("model")
+            if civitai_model and not metadata.get("checkpoint"):
+                metadata["checkpoint"] = civitai_model
+        elif parsed_embedded:
             parsed_loras = parsed_embedded.get("loras")
             if parsed_loras and not metadata.get("loras"):
                 metadata["loras"] = parsed_loras
@@ -1270,16 +1289,29 @@ class RecipeManagementHandler:
 
             with open(temp_path, "rb") as file_obj:
                 model_ver_id = None
+                civitai_meta_raw = (
+                    image_info.get("meta") if civitai_image_id and image_info else None
+                )
                 if civitai_image_id and image_info:
                     model_ver_id = image_info.get("modelVersionId")
                     if not model_ver_id:
                         ids = image_info.get("modelVersionIds")
                         if isinstance(ids, list) and ids:
                             model_ver_id = ids[0]
+
+                    # Inject root-level modelVersionIds into meta so downstream
+                    # parsers (CivitaiApiMetadataParser) can discover ALL resources
+                    # (checkpoint + LoRAs), not just the first model version ID.
+                    # CivitAI API returns modelVersionIds at the root level of
+                    # the image response, NOT inside the meta object.
+                    mvids = image_info.get("modelVersionIds")
+                    if mvids and isinstance(civitai_meta_raw, dict):
+                        civitai_meta_raw["modelVersionIds"] = mvids
+
                 return (
                     file_obj.read(),
                     extension,
-                    image_info.get("meta") if civitai_image_id and image_info else None,
+                    civitai_meta_raw,
                     model_ver_id,
                 )
         except RecipeDownloadError:
@@ -1467,20 +1499,34 @@ class RecipeManagementHandler:
                 "Failed to extract embedded metadata: %s", exc
             )
 
-        if parsed_embedded is None and civitai_meta_raw:
+        # Parse CivitAI API meta to discover all resources from modelVersionIds.
+        # Run unconditionally — EXIF parsing succeeds for gen_params but misses
+        # LoRAs (modelVersionIds is NOT in the image EXIF).
+        civitai_parsed = None
+        if civitai_meta_raw:
             civitai_inner_meta = civitai_meta_raw
             if isinstance(civitai_meta_raw, dict) and "meta" in civitai_meta_raw:
                 civitai_inner_meta = civitai_meta_raw["meta"]
+                # Propagate modelVersionIds into unwrapped meta — it lives
+                # at the outer meta level in the CivitAI API response.
+                _mvids = civitai_meta_raw.get("modelVersionIds")
+                if _mvids and isinstance(civitai_inner_meta, dict):
+                    civitai_inner_meta["modelVersionIds"] = _mvids
             if isinstance(civitai_inner_meta, dict):
                 parser = self._analysis_service._recipe_parser_factory.create_parser(
                     civitai_inner_meta
                 )
                 if parser:
-                    parsed_embedded = await parser.parse_metadata(
+                    civitai_parsed = await parser.parse_metadata(
                         civitai_inner_meta, recipe_scanner=recipe_scanner
                     )
-                    if parsed_embedded and "gen_params" in parsed_embedded:
+                    if civitai_parsed and "gen_params" in civitai_parsed:
-                        embedded_gen_params = parsed_embedded["gen_params"]
+                        # Merge: API gen_params override EXIF at field level,
+                        # EXIF fills in fields the API doesn't have.
+                        embedded_gen_params = {
+                            **(embedded_gen_params or {}),
+                            **civitai_parsed["gen_params"],
+                        }
 
         metadata: Dict[str, Any] = {
             "base_model": "",
@@ -1489,7 +1535,14 @@ class RecipeManagementHandler:
             "source_path": image_url,
         }
 
-        if parsed_embedded:
+        if civitai_parsed:
+            civitai_loras = civitai_parsed.get("loras", [])
+            if civitai_loras and not metadata.get("loras"):
+                metadata["loras"] = civitai_loras
+            civitai_model = civitai_parsed.get("model")
+            if civitai_model and not metadata.get("checkpoint"):
+                metadata["checkpoint"] = civitai_model
+        elif parsed_embedded:
             parsed_loras = parsed_embedded.get("loras")
             if parsed_loras and not metadata.get("loras"):
                 metadata["loras"] = parsed_loras

tests/routes/test_recipe_routes.py

@@ -785,10 +785,16 @@ async def test_import_remote_recipe_merges_metadata(
         async def parse_metadata(self, raw, recipe_scanner=None):
             return json.loads(raw[len("Recipe metadata: ") :])
 
+    class MockApiParser:
+        async def parse_metadata(self, raw, recipe_scanner=None):
+            return {"gen_params": raw, "loras": []}
+
     class MockFactory:
         def create_parser(self, raw):
-            if raw.startswith("Recipe metadata: "):
+            if isinstance(raw, str) and raw.startswith("Recipe metadata: "):
                 return MockParser()
+            if isinstance(raw, dict):
+                return MockApiParser()
             return None
 
     # 4. Setup Harness and run test

tests/services/test_civitai_client.py

@@ -222,7 +222,7 @@ async def test_get_model_versions_raises_on_other_errors(monkeypatch, downloader
 async def test_get_model_versions_bulk_success(monkeypatch, downloader):
     async def fake_make_request(method, url, use_auth=True, **kwargs):
         assert url.endswith("/models")
-        assert kwargs.get("params") == {"ids": "1,2"}
+        assert kwargs.get("params") == {"ids": "1,2", "nsfw": "true"}
         return True, {
             "items": [
                 {