fix: Fix duplicate file blocks (#3105)

letta/schemas/memory.py

@@ -1,7 +1,8 @@
+import logging
 from typing import TYPE_CHECKING, List, Optional
 
 from jinja2 import Template, TemplateSyntaxError
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, field_validator
 
 # Forward referencing to avoid circular import with Agent -> Memory -> Agent
 if TYPE_CHECKING:
@@ -69,6 +70,30 @@ class Memory(BaseModel, validate_assignment=True):
         default_factory=list, description="Blocks representing the agent's in-context memory of an attached file"
     )
 
+    @field_validator("file_blocks")
+    @classmethod
+    def validate_file_blocks_no_duplicates(cls, v: List[Block]) -> List[Block]:
+        """Validate that file_blocks don't contain duplicate labels, log warnings and remove duplicates."""
+        if not v:
+            return v
+
+        seen_labels = set()
+        unique_blocks = []
+        duplicate_labels = []
+
+        for block in v:
+            if block.label in seen_labels:
+                duplicate_labels.append(block.label)
+            else:
+                seen_labels.add(block.label)
+                unique_blocks.append(block)
+
+        if duplicate_labels:
+            logger = logging.getLogger(__name__)
+            logger.warning(f"Duplicate block labels found in file_blocks: {duplicate_labels}. Removing duplicates.")
+
+        return unique_blocks
+
     # Memory.template is a Jinja2 template for compiling memory module into a prompt string.
     prompt_template: str = Field(
         default="{% for block in blocks %}"

letta/services/agent_manager.py

@@ -1730,7 +1730,9 @@ class AgentManager:
             agent_state.memory.blocks = [b for b in blocks if b is not None]
 
         if file_block_names:
-            file_blocks = await self.file_agent_manager.get_all_file_blocks_by_name(file_names=file_block_names, actor=actor)
+            file_blocks = await self.file_agent_manager.get_all_file_blocks_by_name(
+                file_names=file_block_names, agent_id=agent_state.id, actor=actor
+            )
             agent_state.memory.file_blocks = [b for b in file_blocks if b is not None]
 
         return agent_state

letta/services/files_agents_manager.py

@@ -165,17 +165,19 @@ class FileAgentManager:
         self,
         *,
         file_names: List[str],
+        agent_id: str,
         actor: PydanticUser,
     ) -> List[PydanticBlock]:
         """
-        Retrieve multiple FileAgent associations by their IDs in a single query.
+        Retrieve multiple FileAgent associations by their file names for a specific agent.
 
         Args:
             file_names: List of file names to retrieve
+            agent_id: ID of the agent to retrieve file blocks for
             actor: The user making the request
 
         Returns:
-            List of PydanticFileAgent objects found (may be fewer than requested if some IDs don't exist)
+            List of PydanticBlock objects found (may be fewer than requested if some file names don't exist)
         """
         if not file_names:
             return []
@@ -185,6 +187,7 @@ class FileAgentManager:
             query = select(FileAgentModel).where(
                 and_(
                     FileAgentModel.file_name.in_(file_names),
+                    FileAgentModel.agent_id == agent_id,
                     FileAgentModel.organization_id == actor.organization_id,
                 )
             )