refactor(vector_store): update memory operations to use async search

- Update load_today_memory_op.py to use async version of _retrieve_today_memories
- Update memory_deduplication_op.py to use async version of _deduplicate_task_memories and _get_existing_task_memory_embeddings
- Update retrieve_memory_op.py to use async_search instead of search

reme_ai/retrieve/personal/retrieve_memory_op.py

@@ -23,7 +23,7 @@ class RetrieveMemoryOp(BaseLLMOp):
         assert query, "query should be not empty!"
 
         workspace_id: str = self.context.workspace_id
-        nodes: List[VectorNode] = self.vector_store.search(query=query, workspace_id=workspace_id, top_k=top_k)
+        nodes: List[VectorNode] = await self.vector_store.async_search(query=query, workspace_id=workspace_id, top_k=top_k)
         memory_list: List[BaseMemory] = []
         memory_content_list: List[str] = []
         for node in nodes:

reme_ai/summary/personal/load_today_memory_op.py

@@ -36,7 +36,7 @@ class LoadTodayMemoryOp(BaseLLMOp):
         logger.info(f"Loading today's memories for user: {user_name} (top_k: {top_k})")
 
         # Get today's memories from vector store
-        today_memories = self._retrieve_today_memories(workspace_id, user_name, top_k)
+        today_memories = await self._retrieve_today_memories(workspace_id, user_name, top_k)
 
         if not today_memories:
             logger.info("No memories found for today")
@@ -47,7 +47,7 @@ class LoadTodayMemoryOp(BaseLLMOp):
         self.context.today_memories = today_memories
         logger.info(f"Final today's memory list size: {len(today_memories)}")
 
-    def _retrieve_today_memories(self, workspace_id: str, user_name: str, top_k: int) -> List[BaseMemory]:
+    async def _retrieve_today_memories(self, workspace_id: str, user_name: str, top_k: int) -> List[BaseMemory]:
         """
         Retrieve memories from today using vector store with date filtering.
         
@@ -74,7 +74,7 @@ class LoadTodayMemoryOp(BaseLLMOp):
             }
 
             # Search vector store with date filter
-            nodes: List[VectorNode] = self.vector_store.search(
+            nodes: List[VectorNode] = await self.vector_store.async_search(
                 query="",  # Empty query to get all results for today
                 workspace_id=workspace_id,
                 top_k=top_k,

reme_ai/summary/task/memory_deduplication_op.py

@@ -22,7 +22,7 @@ class MemoryDeduplicationOp(BaseOp):
         logger.info(f"Starting deduplication for {len(task_memories)} task memories")
 
         # Perform deduplication
-        deduplicated_task_memories = self._deduplicate_task_memories(task_memories)
+        deduplicated_task_memories = await self._deduplicate_task_memories(task_memories)
 
         logger.info(
             f"Deduplication complete: {len(deduplicated_task_memories)} deduplicated task memories out of {len(task_memories)}")
@@ -30,7 +30,7 @@ class MemoryDeduplicationOp(BaseOp):
         # Update context
         self.context.memory_list = deduplicated_task_memories
 
-    def _deduplicate_task_memories(self, task_memories: List[BaseMemory]) -> List[BaseMemory]:
+    async def _deduplicate_task_memories(self, task_memories: List[BaseMemory]) -> List[BaseMemory]:
         """Remove duplicate task memories"""
         if not task_memories:
             return task_memories
@@ -41,7 +41,7 @@ class MemoryDeduplicationOp(BaseOp):
         unique_task_memories = []
 
         # Get existing task memory embeddings
-        existing_embeddings = self._get_existing_task_memory_embeddings(workspace_id)
+        existing_embeddings = await self._get_existing_task_memory_embeddings(workspace_id)
 
         for task_memory in task_memories:
             # Generate embedding for current task memory
@@ -67,14 +67,14 @@ class MemoryDeduplicationOp(BaseOp):
 
         return unique_task_memories
 
-    def _get_existing_task_memory_embeddings(self, workspace_id: str) -> List[List[float]]:
+    async def _get_existing_task_memory_embeddings(self, workspace_id: str) -> List[List[float]]:
         """Get embeddings of existing task memories"""
         try:
             if not hasattr(self.context, 'vector_store') or not self.context.vector_store or not workspace_id:
                 return []
 
             # Query existing task memory nodes
-            existing_nodes = self.context.vector_store.search(
+            existing_nodes = await self.context.vector_store.async_search(
                 query="...",  # Empty query to get all
                 workspace_id=workspace_id,
                 top_k=self.op_params.get("max_existing_task_memories", 1000)