fix: support custom model architectures and restore model param wiring in search

code_review_graph/embeddings.py

@@ -62,7 +62,11 @@ def _get_model(self):
         if self._model is None:
             try:
                 from sentence_transformers import SentenceTransformer
-                self._model = SentenceTransformer(self._model_name)
+                self._model = SentenceTransformer(
+                    self._model_name,
+                    trust_remote_code=True,
+                    model_kwargs={"trust_remote_code": True},
+                )
             except ImportError:
                 raise ImportError(
                     "sentence-transformers not installed. "

code_review_graph/search.py

@@ -168,6 +168,7 @@ def _embedding_search(
     store: GraphStore,
     query: str,
     limit: int = 50,
+    model: str | None = None,
 ) -> list[tuple[int, float]]:
     """Run a vector similarity search using the embedding store.
 
@@ -180,7 +181,7 @@ def _embedding_search(
         return []
 
     try:
-        emb_store = EmbeddingStore(store.db_path)
+        emb_store = EmbeddingStore(store.db_path, model=model)
         try:
             if not emb_store.available or emb_store.count() == 0:
                 return []
@@ -264,6 +265,7 @@ def hybrid_search(
     kind: Optional[str] = None,
     limit: int = 20,
     context_files: Optional[list[str]] = None,
+    model: Optional[str] = None,
 ) -> list[dict[str, Any]]:
     """Hybrid search combining FTS5 BM25 and vector embeddings via RRF.
 
@@ -301,7 +303,7 @@ def hybrid_search(
         logger.warning("FTS5 unavailable, will use fallback: %s", e)
 
     # Try embedding search
-    emb_results = _embedding_search(store, query, limit=fetch_limit)
+    emb_results = _embedding_search(store, query, limit=fetch_limit, model=model)
 
     # ------ Phase 2: Merge via RRF or fallback ------
     if fts_results or emb_results:

code_review_graph/tools/query.py

@@ -321,6 +321,7 @@ def semantic_search_nodes(
     try:
         results = hybrid_search(
             store, query, kind=kind, limit=limit, context_files=context_files,
+            model=model,
         )
 
         search_mode = "hybrid"