list_memories

Instructions

Implementation Reference

• mcp_mitm_mem0/mcp_server.py:217-278 (handler)

  The handler function for the 'list_memories' MCP tool. Registered via @mcp.tool decorator. Accepts optional user_id, fetches all memories via memory_service, includes logging and error handling.

  @mcp.tool(name="list_memories", description="List all stored conversation memories")
  async def list_memories(user_id: str | None = None) -> list[dict[str, Any]]:
      """
      List all memories for a user, providing a complete conversation history.
  
      ## When to Use
  
      - User wants to see their complete conversation history
      - You need to browse through all memories
      - Performing a comprehensive review or audit
      - Search didn't find specific memories
      - Getting an overview of all stored conversations
  
      ## Example Usage
  
      ```python
      # List all memories for default user
      memories = await list_memories()
  
      # List memories for specific user
      memories = await list_memories(user_id="alice@example.com")
      ```
  
      ## Example Response
  
      ```json
      [
          {
              "id": "mem_xyz789",
              "memory": "Discussion about implementing OAuth2 with refresh tokens...",
              "created_at": "2024-01-20T14:30:00Z",
              "metadata": {"type": "conversation", "topic": "authentication"}
          },
          {
              "id": "mem_abc123",
              "memory": "User mentioned preferring PostgreSQL over MySQL...",
              "created_at": "2024-01-19T09:15:00Z",
              "metadata": {"type": "preference"}
          }
      ]
      ```
  
      Note: This returns ALL memories. For large histories, prefer search_memories() for specific topics.
  
      Args:
          user_id: User ID (optional, defaults to DEFAULT_USER_ID from settings)
  
      Returns:
          List of all memories sorted by creation date (newest first), each containing:
          - id: Unique memory identifier
          - memory/content: The stored conversation text
          - created_at: ISO timestamp of memory creation
          - metadata: Additional context
      """
      try:
          results = await memory_service.get_all_memories(user_id=user_id)
          logger.info("Memory list retrieved", memory_count=len(results))
          return results
      except Exception as e:
          logger.error("List failed", error=str(e))
          raise RuntimeError(f"List failed: {str(e)}") from e

• mcp_mitm_mem0/memory_service.py:189-215 (helper)

  The MemoryService.get_all_memories method called by the tool handler. Uses Mem0's AsyncMemoryClient to retrieve all memories for the given user_id.

  async def get_all_memories(
      self, user_id: str | None = None
  ) -> list[dict[str, Any]]:
      """Get all memories for a user asynchronously.
  
      Args:
          user_id: User identifier (defaults to settings.default_user_id)
  
      Returns:
          List of all memories for the user
      """
      user_id = user_id or settings.default_user_id
  
      try:
          self._logger.info("Getting all memories", user_id=user_id)
  
          results = await self.async_client.get_all(user_id=user_id, version="v2")
  
          self._logger.info(
              "Retrieved memories", user_id=user_id, memory_count=len(results)
          )
          return results
  
      except Exception as e:
          self._logger.error("Failed to get memories", user_id=user_id, error=str(e))
          raise

• mcp_mitm_mem0/mcp_server.py:217-217 (registration)

  The @mcp.tool decorator that registers the list_memories function as an MCP tool.

  @mcp.tool(name="list_memories", description="List all stored conversation memories")