agno 2.1.0__tar.gz → 2.1.2__tar.gz

{agno-2.1.0 → agno-2.1.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agno
-Version: 2.1.0
+Version: 2.1.2
 Summary: Agno: a lightweight library for building Multi-Agent Systems
 Author-email: Ashpreet Bedi <ashpreet@agno.com>
 Project-URL: homepage, https://agno.com
@@ -23,6 +23,7 @@ Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: docstring-parser
 Requires-Dist: gitpython
+Requires-Dist: h11>=0.16.0
 Requires-Dist: httpx
 Requires-Dist: packaging
 Requires-Dist: pydantic-settings
@@ -247,6 +248,8 @@ Requires-Dist: chonkie[st]; extra == "chonkie"
 Requires-Dist: chonkie; extra == "chonkie"
 Provides-Extra: agui
 Requires-Dist: ag-ui-protocol; extra == "agui"
+Provides-Extra: a2a
+Requires-Dist: a2a-sdk; extra == "a2a"
 Provides-Extra: huggingface
 Requires-Dist: huggingface-hub; extra == "huggingface"
 Provides-Extra: performance
@@ -267,7 +270,6 @@ Requires-Dist: agno[anthropic]; extra == "models"
 Requires-Dist: agno[azure]; extra == "models"
 Requires-Dist: agno[cerebras]; extra == "models"
 Requires-Dist: agno[cohere]; extra == "models"
-Requires-Dist: agno[infinity]; extra == "models"
 Requires-Dist: agno[google]; extra == "models"
 Requires-Dist: agno[groq]; extra == "models"
 Requires-Dist: agno[ibm]; extra == "models"
@@ -379,11 +381,11 @@ Dynamic: license-file
 
 ## What is Agno?
 
-[Agno](https://docs.agno.com) is a high-performance runtime for multi-agent systems. Use it to build, run and manage secure multi-agent systems in your cloud.
+[Agno](https://docs.agno.com) is a high-performance SDK and runtime for multi-agent systems. Use it to build, run and manage multi-agent systems in your cloud.
 
-Agno gives you the fastest framework for building agents with session management, memory, knowledge, human in the loop and MCP support. You can put agents together as an autonomous multi-agent team, or build step-based agentic workflows for full control over complex multi-step processes.
+Agno is the fastest framework for building agents with built-in memory, knowledge, session management, human in the loop and best-in-class MCP support. You can put agents together as multi-agent teams or step-based agentic workflows.
 
-In 10 lines of code, we can build an Agent that will fetch the top stories from HackerNews and summarize them.
+In 10 lines of code, we can build an Agent that uses tools to achieve a task.
 
 ```python hackernews_agent.py
 from agno.agent import Agent
@@ -391,22 +393,22 @@ from agno.models.anthropic import Claude
 from agno.tools.hackernews import HackerNewsTools
 
 agent = Agent(
-    model=Claude(id="claude-sonnet-4-0"),
+    model=Claude(id="claude-sonnet-4-5"),
     tools=[HackerNewsTools()],
     markdown=True,
 )
-agent.print_response("Summarize the top 5 stories on hackernews", stream=True)
+agent.print_response("Write a report on trending startups and products.", stream=True)
 ```
 
 But the real advantage of Agno is its [AgentOS](https://docs.agno.com/agent-os/introduction) runtime:
 
-1. You get a pre-built FastAPI app for running your agentic system, meaning you start building your product on day one. This is a remarkable advantage over other solutions or rolling your own.
-2. You also get a control plane which connects directly to your AgentOS for testing, monitoring and managing your system. This gives you unmatched visibility and control over your system.
-3. Your AgentOS runs in your cloud and you get complete data privacy because no data ever leaves your system. This is incredible for security conscious enterprises that can't send traces to external services.
+1. You get a pre-built FastAPI app for serving your agents, teams and workflows, meaning you start building your AI product on day one. This is a remarkable advantage over other solutions.
+2. You also get a UI that connects directly to the pre-built FastAPI app. Use it to test, monitor and manage your system. This gives you unmatched visibility and control.
+3. Your AgentOS runs in your cloud and you get complete privacy because no data ever leaves your system. This is incredible for security conscious enterprises that can't send data to external services.
 
-For organizations building agents, Agno provides the complete solution. You get the fastest framework for building agents (speed of development and execution), a pre-built FastAPI app that lets you build your product on day one, and a control plane for managing your system.
+For organizations building agents, Agno provides the complete solution. You get the fastest framework for building agents (speed of development and execution), a pre-built FastAPI app that get you building product on day one, and a control plane for managing your system.
 
-We bring a novel architecture that no other framework provides, your AgentOS runs securely in your cloud, and the control plane connects directly to it from your browser. You don't need to send data to external services or pay retention costs, you get complete privacy and control.
+We bring a novel architecture that no other framework provides, your AgentOS runs securely in your cloud, and the control plane connects directly to it from your browser. You don't need to send data to any external services or pay retention costs, you get complete privacy and control.
 
 ## Getting started
 
@@ -463,10 +465,10 @@ source .venvs/perfenv/bin/activate
 # pip install openai agno langgraph langchain_openai
 
 # Agno
-python evals/performance/instantiation_with_tool.py
+python cookbook/evals/performance/instantiate_agent_with_tool.py
 
 # LangGraph
-python evals/performance/other/langgraph_instantiation.py
+python cookbook/evals/performance/comparison/langgraph_instantiation.py
 ```
 
 > The following evaluation is run on an Apple M4 MacBook Pro. It also runs as a Github action on this repo.

{agno-2.1.0 → agno-2.1.2}/README.md

@@ -16,11 +16,11 @@
 
 ## What is Agno?
 
-[Agno](https://docs.agno.com) is a high-performance runtime for multi-agent systems. Use it to build, run and manage secure multi-agent systems in your cloud.
+[Agno](https://docs.agno.com) is a high-performance SDK and runtime for multi-agent systems. Use it to build, run and manage multi-agent systems in your cloud.
 
-Agno gives you the fastest framework for building agents with session management, memory, knowledge, human in the loop and MCP support. You can put agents together as an autonomous multi-agent team, or build step-based agentic workflows for full control over complex multi-step processes.
+Agno is the fastest framework for building agents with built-in memory, knowledge, session management, human in the loop and best-in-class MCP support. You can put agents together as multi-agent teams or step-based agentic workflows.
 
-In 10 lines of code, we can build an Agent that will fetch the top stories from HackerNews and summarize them.
+In 10 lines of code, we can build an Agent that uses tools to achieve a task.
 
 ```python hackernews_agent.py
 from agno.agent import Agent
@@ -28,22 +28,22 @@ from agno.models.anthropic import Claude
 from agno.tools.hackernews import HackerNewsTools
 
 agent = Agent(
-    model=Claude(id="claude-sonnet-4-0"),
+    model=Claude(id="claude-sonnet-4-5"),
     tools=[HackerNewsTools()],
     markdown=True,
 )
-agent.print_response("Summarize the top 5 stories on hackernews", stream=True)
+agent.print_response("Write a report on trending startups and products.", stream=True)
 ```
 
 But the real advantage of Agno is its [AgentOS](https://docs.agno.com/agent-os/introduction) runtime:
 
-1. You get a pre-built FastAPI app for running your agentic system, meaning you start building your product on day one. This is a remarkable advantage over other solutions or rolling your own.
-2. You also get a control plane which connects directly to your AgentOS for testing, monitoring and managing your system. This gives you unmatched visibility and control over your system.
-3. Your AgentOS runs in your cloud and you get complete data privacy because no data ever leaves your system. This is incredible for security conscious enterprises that can't send traces to external services.
+1. You get a pre-built FastAPI app for serving your agents, teams and workflows, meaning you start building your AI product on day one. This is a remarkable advantage over other solutions.
+2. You also get a UI that connects directly to the pre-built FastAPI app. Use it to test, monitor and manage your system. This gives you unmatched visibility and control.
+3. Your AgentOS runs in your cloud and you get complete privacy because no data ever leaves your system. This is incredible for security conscious enterprises that can't send data to external services.
 
-For organizations building agents, Agno provides the complete solution. You get the fastest framework for building agents (speed of development and execution), a pre-built FastAPI app that lets you build your product on day one, and a control plane for managing your system.
+For organizations building agents, Agno provides the complete solution. You get the fastest framework for building agents (speed of development and execution), a pre-built FastAPI app that get you building product on day one, and a control plane for managing your system.
 
-We bring a novel architecture that no other framework provides, your AgentOS runs securely in your cloud, and the control plane connects directly to it from your browser. You don't need to send data to external services or pay retention costs, you get complete privacy and control.
+We bring a novel architecture that no other framework provides, your AgentOS runs securely in your cloud, and the control plane connects directly to it from your browser. You don't need to send data to any external services or pay retention costs, you get complete privacy and control.
 
 ## Getting started
 
@@ -100,10 +100,10 @@ source .venvs/perfenv/bin/activate
 # pip install openai agno langgraph langchain_openai
 
 # Agno
-python evals/performance/instantiation_with_tool.py
+python cookbook/evals/performance/instantiate_agent_with_tool.py
 
 # LangGraph
-python evals/performance/other/langgraph_instantiation.py
+python cookbook/evals/performance/comparison/langgraph_instantiation.py
 ```
 
 > The following evaluation is run on an Apple M4 MacBook Pro. It also runs as a Github action on this repo.

{agno-2.1.0 → agno-2.1.2}/agno/agent/agent.py

@@ -5244,7 +5244,7 @@ class Agent:
 
         # 3.2.5 Add information about agentic filters if enabled
         if self.knowledge is not None and self.enable_agentic_knowledge_filters:
-            valid_filters = getattr(self.knowledge, "valid_metadata_filters", None)
+            valid_filters = self.knowledge.get_valid_filters()
             if valid_filters:
                 valid_filters_str = ", ".join(valid_filters)
                 additional_information.append(
@@ -6420,6 +6420,12 @@ class Agent:
                         log_warning("Reasoning error. Reasoning response is empty, continuing regular session...")
                         break
 
+                    if isinstance(reasoning_agent_response.content, str):
+                        log_warning(
+                            "Reasoning error. Content is a string, not structured output. Continuing regular session..."
+                        )
+                        break
+
                     if (
                         reasoning_agent_response.content.reasoning_steps is None
                         or len(reasoning_agent_response.content.reasoning_steps) == 0
@@ -6649,6 +6655,12 @@ class Agent:
                         log_warning("Reasoning error. Reasoning response is empty, continuing regular session...")
                         break
 
+                    if isinstance(reasoning_agent_response.content, str):
+                        log_warning(
+                            "Reasoning error. Content is a string, not structured output. Continuing regular session..."
+                        )
+                        break
+
                     if reasoning_agent_response.content.reasoning_steps is None:
                         log_warning("Reasoning error. Reasoning steps are empty, continuing regular session...")
                         break

{agno-2.1.0 → agno-2.1.2}/agno/db/base.py

@@ -95,20 +95,23 @@ class BaseDb(ABC):
         raise NotImplementedError
 
     @abstractmethod
-    def delete_user_memory(self, memory_id: str) -> None:
+    def delete_user_memory(self, memory_id: str, user_id: Optional[str] = None) -> None:
         raise NotImplementedError
 
     @abstractmethod
-    def delete_user_memories(self, memory_ids: List[str]) -> None:
+    def delete_user_memories(self, memory_ids: List[str], user_id: Optional[str] = None) -> None:
         raise NotImplementedError
 
     @abstractmethod
-    def get_all_memory_topics(self) -> List[str]:
+    def get_all_memory_topics(self, user_id: Optional[str] = None) -> List[str]:
         raise NotImplementedError
 
     @abstractmethod
     def get_user_memory(
-        self, memory_id: str, deserialize: Optional[bool] = True
+        self,
+        memory_id: str,
+        deserialize: Optional[bool] = True,
+        user_id: Optional[str] = None,
     ) -> Optional[Union[UserMemory, Dict[str, Any]]]:
         raise NotImplementedError
 
@@ -133,6 +136,7 @@ class BaseDb(ABC):
         self,
         limit: Optional[int] = None,
         page: Optional[int] = None,
+        user_id: Optional[str] = None,
     ) -> Tuple[List[Dict[str, Any]], int]:
         raise NotImplementedError
 

{agno-2.1.0 → agno-2.1.2}/agno/db/dynamo/dynamo.py

@@ -562,17 +562,31 @@ class DynamoDb(BaseDb):
 
     # --- User Memory ---
 
-    def delete_user_memory(self, memory_id: str) -> None:
+    def delete_user_memory(self, memory_id: str, user_id: Optional[str] = None) -> None:
         """
         Delete a user memory from the database.
 
         Args:
             memory_id: The ID of the memory to delete.
+            user_id: The ID of the user (optional, for filtering).
 
         Raises:
             Exception: If any error occurs while deleting the user memory.
         """
         try:
+            # If user_id is provided, verify the memory belongs to the user before deleting
+            if user_id:
+                response = self.client.get_item(
+                    TableName=self.memory_table_name,
+                    Key={"memory_id": {"S": memory_id}},
+                )
+                item = response.get("Item")
+                if item:
+                    memory_data = deserialize_from_dynamodb_item(item)
+                    if memory_data.get("user_id") != user_id:
+                        log_debug(f"Memory {memory_id} does not belong to user {user_id}")
+                        return
+
             self.client.delete_item(
                 TableName=self.memory_table_name,
                 Key={"memory_id": {"S": memory_id}},
@@ -583,18 +597,34 @@ class DynamoDb(BaseDb):
             log_error(f"Failed to delete user memory {memory_id}: {e}")
             raise e
 
-    def delete_user_memories(self, memory_ids: List[str]) -> None:
+    def delete_user_memories(self, memory_ids: List[str], user_id: Optional[str] = None) -> None:
         """
         Delete user memories from the database in batches.
 
         Args:
             memory_ids: List of memory IDs to delete
+            user_id: The ID of the user (optional, for filtering).
 
         Raises:
             Exception: If any error occurs while deleting the user memories.
         """
 
         try:
+            # If user_id is provided, filter memory_ids to only those belonging to the user
+            if user_id:
+                filtered_memory_ids = []
+                for memory_id in memory_ids:
+                    response = self.client.get_item(
+                        TableName=self.memory_table_name,
+                        Key={"memory_id": {"S": memory_id}},
+                    )
+                    item = response.get("Item")
+                    if item:
+                        memory_data = deserialize_from_dynamodb_item(item)
+                        if memory_data.get("user_id") == user_id:
+                            filtered_memory_ids.append(memory_id)
+                memory_ids = filtered_memory_ids
+
             for i in range(0, len(memory_ids), DYNAMO_BATCH_SIZE_LIMIT):
                 batch = memory_ids[i : i + DYNAMO_BATCH_SIZE_LIMIT]
 
@@ -611,6 +641,9 @@ class DynamoDb(BaseDb):
     def get_all_memory_topics(self) -> List[str]:
         """Get all memory topics from the database.
 
+        Args:
+            user_id: The ID of the user (optional, for filtering).
+
         Returns:
             List[str]: List of unique memory topics.
         """
@@ -619,13 +652,17 @@ class DynamoDb(BaseDb):
             if table_name is None:
                 return []
 
-            # Scan the entire table to get all memories
-            response = self.client.scan(TableName=table_name)
+            # Build filter expression for user_id if provided
+            scan_kwargs = {"TableName": table_name}
+
+            # Scan the table to get memories
+            response = self.client.scan(**scan_kwargs)
             items = response.get("Items", [])
 
             # Handle pagination
             while "LastEvaluatedKey" in response:
-                response = self.client.scan(TableName=table_name, ExclusiveStartKey=response["LastEvaluatedKey"])
+                scan_kwargs["ExclusiveStartKey"] = response["LastEvaluatedKey"]
+                response = self.client.scan(**scan_kwargs)
                 items.extend(response.get("Items", []))
 
             # Extract topics from all memories
@@ -642,13 +679,18 @@ class DynamoDb(BaseDb):
             raise e
 
     def get_user_memory(
-        self, memory_id: str, deserialize: Optional[bool] = True
+        self,
+        memory_id: str,
+        deserialize: Optional[bool] = True,
+        user_id: Optional[str] = None,
     ) -> Optional[Union[UserMemory, Dict[str, Any]]]:
         """
         Get a user memory from the database as a UserMemory object.
 
         Args:
             memory_id: The ID of the memory to get.
+            deserialize: Whether to deserialize the memory.
+            user_id: The ID of the user (optional, for filtering).
 
         Returns:
             Optional[UserMemory]: The user memory data if found, None otherwise.
@@ -665,6 +707,11 @@ class DynamoDb(BaseDb):
                 return None
 
             item = deserialize_from_dynamodb_item(item)
+
+            # Filter by user_id if provided
+            if user_id and item.get("user_id") != user_id:
+                return None
+
             if not deserialize:
                 return item
 
@@ -804,6 +851,7 @@ class DynamoDb(BaseDb):
         Args:
             limit (Optional[int]): The maximum number of user stats to return.
             page (Optional[int]): The page number.
+            user_id (Optional[str]): The ID of the user (optional, for filtering).
 
         Returns:
             Tuple[List[Dict[str, Any]], int]: A list of dictionaries containing user stats and total count.
@@ -823,29 +871,33 @@ class DynamoDb(BaseDb):
         try:
             table_name = self._get_table("memories")
 
-            response = self.client.scan(TableName=table_name)
+            # Build filter expression for user_id if provided
+            scan_kwargs = {"TableName": table_name}
+
+            response = self.client.scan(**scan_kwargs)
             items = response.get("Items", [])
 
             # Handle pagination
             while "LastEvaluatedKey" in response:
-                response = self.client.scan(TableName=table_name, ExclusiveStartKey=response["LastEvaluatedKey"])
+                scan_kwargs["ExclusiveStartKey"] = response["LastEvaluatedKey"]
+                response = self.client.scan(**scan_kwargs)
                 items.extend(response.get("Items", []))
 
             # Aggregate stats by user_id
             user_stats = {}
             for item in items:
                 memory_data = deserialize_from_dynamodb_item(item)
-                user_id = memory_data.get("user_id")
+                current_user_id = memory_data.get("user_id")
 
-                if user_id:
-                    if user_id not in user_stats:
-                        user_stats[user_id] = {
-                            "user_id": user_id,
+                if current_user_id:
+                    if current_user_id not in user_stats:
+                        user_stats[current_user_id] = {
+                            "user_id": current_user_id,
                             "total_memories": 0,
                             "last_memory_updated_at": None,
                         }
 
-                    user_stats[user_id]["total_memories"] += 1
+                    user_stats[current_user_id]["total_memories"] += 1
 
                     updated_at = memory_data.get("updated_at")
                     if updated_at:
@@ -853,10 +905,10 @@ class DynamoDb(BaseDb):
                         updated_at_timestamp = int(updated_at_dt.timestamp())
 
                         if updated_at_timestamp and (
-                            user_stats[user_id]["last_memory_updated_at"] is None
-                            or updated_at_timestamp > user_stats[user_id]["last_memory_updated_at"]
+                            user_stats[current_user_id]["last_memory_updated_at"] is None
+                            or updated_at_timestamp > user_stats[current_user_id]["last_memory_updated_at"]
                         ):
-                            user_stats[user_id]["last_memory_updated_at"] = updated_at_timestamp
+                            user_stats[current_user_id]["last_memory_updated_at"] = updated_at_timestamp
 
             # Convert to list and apply sorting
             stats_list = list(user_stats.values())

{agno-2.1.0 → agno-2.1.2}/agno/db/firestore/firestore.py

@@ -596,11 +596,12 @@ class FirestoreDb(BaseDb):
 
     # -- Memory methods --
 
-    def delete_user_memory(self, memory_id: str):
+    def delete_user_memory(self, memory_id: str, user_id: Optional[str] = None):
         """Delete a user memory from the database.
 
         Args:
             memory_id (str): The ID of the memory to delete.
+            user_id (Optional[str]): The ID of the user (optional, for filtering).
 
         Returns:
             bool: True if the memory was deleted, False otherwise.
@@ -610,28 +611,41 @@ class FirestoreDb(BaseDb):
         """
         try:
             collection_ref = self._get_collection(table_type="memories")
-            docs = collection_ref.where(filter=FieldFilter("memory_id", "==", memory_id)).stream()
 
-            deleted_count = 0
-            for doc in docs:
-                doc.reference.delete()
-                deleted_count += 1
-
-            success = deleted_count > 0
-            if success:
-                log_debug(f"Successfully deleted user memory id: {memory_id}")
+            # If user_id is provided, verify the memory belongs to the user before deleting
+            if user_id:
+                docs = collection_ref.where(filter=FieldFilter("memory_id", "==", memory_id)).stream()
+                for doc in docs:
+                    data = doc.to_dict()
+                    if data.get("user_id") != user_id:
+                        log_debug(f"Memory {memory_id} does not belong to user {user_id}")
+                        return
+                    doc.reference.delete()
+                    log_debug(f"Successfully deleted user memory id: {memory_id}")
+                    return
             else:
-                log_debug(f"No user memory found with id: {memory_id}")
+                docs = collection_ref.where(filter=FieldFilter("memory_id", "==", memory_id)).stream()
+                deleted_count = 0
+                for doc in docs:
+                    doc.reference.delete()
+                    deleted_count += 1
+
+                success = deleted_count > 0
+                if success:
+                    log_debug(f"Successfully deleted user memory id: {memory_id}")
+                else:
+                    log_debug(f"No user memory found with id: {memory_id}")
 
         except Exception as e:
             log_error(f"Error deleting user memory: {e}")
             raise e
 
-    def delete_user_memories(self, memory_ids: List[str]) -> None:
+    def delete_user_memories(self, memory_ids: List[str], user_id: Optional[str] = None) -> None:
         """Delete user memories from the database.
 
         Args:
             memory_ids (List[str]): The IDs of the memories to delete.
+            user_id (Optional[str]): The ID of the user (optional, for filtering).
 
         Raises:
             Exception: If there is an error deleting the memories.
@@ -641,11 +655,21 @@ class FirestoreDb(BaseDb):
             batch = self.db_client.batch()
             deleted_count = 0
 
-            for memory_id in memory_ids:
-                docs = collection_ref.where(filter=FieldFilter("memory_id", "==", memory_id)).stream()
-                for doc in docs:
-                    batch.delete(doc.reference)
-                    deleted_count += 1
+            # If user_id is provided, filter memory_ids to only those belonging to the user
+            if user_id:
+                for memory_id in memory_ids:
+                    docs = collection_ref.where(filter=FieldFilter("memory_id", "==", memory_id)).stream()
+                    for doc in docs:
+                        data = doc.to_dict()
+                        if data.get("user_id") == user_id:
+                            batch.delete(doc.reference)
+                            deleted_count += 1
+            else:
+                for memory_id in memory_ids:
+                    docs = collection_ref.where(filter=FieldFilter("memory_id", "==", memory_id)).stream()
+                    for doc in docs:
+                        batch.delete(doc.reference)
+                        deleted_count += 1
 
             batch.commit()
 
@@ -658,7 +682,9 @@ class FirestoreDb(BaseDb):
             log_error(f"Error deleting memories: {e}")
             raise e
 
-    def get_all_memory_topics(self, create_collection_if_not_found: Optional[bool] = True) -> List[str]:
+    def get_all_memory_topics(
+        self, create_collection_if_not_found: Optional[bool] = True
+    ) -> List[str]:
         """Get all memory topics from the database.
 
         Returns:
@@ -687,12 +713,15 @@ class FirestoreDb(BaseDb):
             log_error(f"Exception getting all memory topics: {e}")
             raise e
 
-    def get_user_memory(self, memory_id: str, deserialize: Optional[bool] = True) -> Optional[UserMemory]:
+    def get_user_memory(
+        self, memory_id: str, deserialize: Optional[bool] = True, user_id: Optional[str] = None
+    ) -> Optional[UserMemory]:
         """Get a memory from the database.
 
         Args:
             memory_id (str): The ID of the memory to get.
             deserialize (Optional[bool]): Whether to serialize the memory. Defaults to True.
+            user_id (Optional[str]): The ID of the user (optional, for filtering).
 
         Returns:
             Optional[UserMemory]:
@@ -711,7 +740,14 @@ class FirestoreDb(BaseDb):
                 result = doc.to_dict()
                 break
 
-            if result is None or not deserialize:
+            if result is None:
+                return None
+
+            # Filter by user_id if provided
+            if user_id and result.get("user_id") != user_id:
+                return None
+
+            if not deserialize:
                 return result
 
             return UserMemory.from_dict(result)
@@ -818,23 +854,26 @@ class FirestoreDb(BaseDb):
         """
         try:
             collection_ref = self._get_collection(table_type="memories")
-            docs = collection_ref.where(filter=FieldFilter("user_id", "!=", None)).stream()
+
+            query = collection_ref.where(filter=FieldFilter("user_id", "!=", None))
+
+            docs = query.stream()
 
             user_stats = {}
             for doc in docs:
                 data = doc.to_dict()
-                user_id = data.get("user_id")
-                if user_id:
-                    if user_id not in user_stats:
-                        user_stats[user_id] = {
-                            "user_id": user_id,
+                current_user_id = data.get("user_id")
+                if current_user_id:
+                    if current_user_id not in user_stats:
+                        user_stats[current_user_id] = {
+                            "user_id": current_user_id,
                             "total_memories": 0,
                             "last_memory_updated_at": 0,
                         }
-                    user_stats[user_id]["total_memories"] += 1
+                    user_stats[current_user_id]["total_memories"] += 1
                     updated_at = data.get("updated_at", 0)
-                    if updated_at > user_stats[user_id]["last_memory_updated_at"]:
-                        user_stats[user_id]["last_memory_updated_at"] = updated_at
+                    if updated_at > user_stats[current_user_id]["last_memory_updated_at"]:
+                        user_stats[current_user_id]["last_memory_updated_at"] = updated_at
 
             # Convert to list and sort
             formatted_results = list(user_stats.values())