diff --git a/.gitignore b/.gitignore
index e9a9653..c743115 100644
--- a/.gitignore
+++ b/.gitignore
@@ -9,4 +9,8 @@ config.yml
 
 # pyc/pyo
 **/*.pyc
-**/*.pyo
\ No newline at end of file
+**/*.pyo
+
+# data
+/data/
+```
\ No newline at end of file
diff --git a/main.py b/main.py
index b990d8d..28e8d51 100644
--- a/main.py
+++ b/main.py
@@ -1,6 +1,9 @@
 from src.liteyuki import *
+from pymongo import MongoClient
+
+a = MongoClient("mongodb://localhost:27017/")
 
 if __name__ == '__main__':
     liteyuki = Liteyuki()
     app = liteyuki.get_asgi()
-    liteyuki.run(app="main:app")
\ No newline at end of file
+    liteyuki.run(app="main:app")
diff --git a/requirements.txt b/requirements.txt
index 60bf44e..32f13d6 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1 +1,3 @@
-nonebot2[fastapi]
\ No newline at end of file
+nonebot2[fastapi]
+keyvalue-sqlite=1.0.7
+motor=3.3.2
\ No newline at end of file
diff --git a/src/api/__init__.py b/src/api/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/src/api/data.py b/src/api/data.py
index e69de29..e9994ec 100644
--- a/src/api/data.py
+++ b/src/api/data.py
@@ -0,0 +1,241 @@
+import sqlite3
+import uuid
+from typing import Any
+
+from pymongo import MongoClient
+from pydantic import BaseModel
+
+
+class LiteModel(BaseModel):
+    pass
+
+
+class BaseORM:
+
+    def __init__(self, *args, **kwargs):
+        pass
+
+    def auto_migrate(self, *args, **kwargs):
+        """自动迁移数据库
+        Args:
+            *args:
+            **kwargs:
+
+        Returns:
+        """
+        raise NotImplementedError
+
+    def save(self, *args, **kwargs):
+        """创建数据
+        Args:
+            *args:
+            **kwargs:
+
+        Returns:
+        """
+        raise NotImplementedError
+
+    def update(self, *args, **kwargs):
+        """更新数据
+        Args:
+            *args:
+            **kwargs:
+
+        Returns:
+        """
+        raise NotImplementedError
+
+    def delete(self, *args, **kwargs):
+        """删除数据
+        Args:
+            *args:
+            **kwargs:
+
+        Returns:
+        """
+        raise NotImplementedError
+
+    def first(self, *args, **kwargs):
+        """查询第一条数据
+        Args:
+            *args:
+            **kwargs:
+
+        Returns:
+        """
+        raise NotImplementedError
+
+    def where(self, *args, **kwargs):
+        """查询数据
+        Args:
+            *args:
+            **kwargs:
+
+        Returns:
+        """
+        raise NotImplementedError
+
+    def all(self, *args, **kwargs):
+        """查询所有数据
+        Args:
+            *args:
+            **kwargs:
+
+        Returns:
+        """
+        raise NotImplementedError
+
+
+class SqliteORM(BaseORM):
+    """同步sqlite数据库操作"""
+    type_map = {
+        int: 'INTEGER',
+        float: 'REAL',
+        str: 'TEXT',
+        bool: 'INTEGER',
+    }
+
+    def __init__(self, db, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        self.db = sqlite3.connect(db)
+
+    @staticmethod
+    def get_model_table_name(model: type(LiteModel) | LiteModel | str) -> str:
+        """获取模型对应的表名"""
+        if isinstance(model, str):
+            return model
+        elif isinstance(model, LiteModel):
+            return model.__class__.__name__
+        elif isinstance(model, type(LiteModel)):
+            return model.__name__
+
+    def auto_migrate(self, *args: type(LiteModel) | LiteModel | str, **kwargs):
+        """自动迁移数据库
+        Args:
+            *args: BaseModel
+            **kwargs:
+                delete_old_columns: bool = False    # 是否删除旧字段
+                add_new_columns: bool = True    # 添加新字段
+        Returns:
+        """
+        for model in args:
+            # 获取模型对应的表名
+            table_name = self.get_model_table_name(model)
+
+            # 获取表中已有的字段
+            existing_columns = set()
+            cursor = self.db.execute(f"PRAGMA table_info({table_name})")
+            for column_info in cursor.fetchall():
+                existing_columns.add(column_info[1])
+
+            # 创建表，如果不存在的话
+            self.db.execute(f'CREATE TABLE IF NOT EXISTS {table_name} (id INTEGER PRIMARY KEY AUTOINCREMENT)')
+
+            # 检测模型中的字段并添加新字段，按照类型添加
+            for field_name, field_type in model.__annotations__.items():
+                if field_name not in existing_columns:
+                    self.db.execute(f'ALTER TABLE {table_name} ADD COLUMN {field_name} {self.type_map.get(field_type, "TEXT")}')
+
+            # 提交事务
+            self.db.commit()
+
+    def save(self, model: LiteModel) -> int:
+        """保存或创建数据，对嵌套模型扁平化处理，加特殊前缀表示为模型，实际储存模型id，嵌套对象单独储存
+        Args:
+            model: BaseModel
+        Returns: id主键
+        """
+        # 先检测表是否存在，不存在则创建
+        table_name = self.get_model_table_name(model)
+        self.db.execute(f'CREATE TABLE IF NOT EXISTS {table_name} (id INTEGER PRIMARY KEY AUTOINCREMENT)')
+
+        # 构建插入语句
+        column_list = []
+        value_list = []
+        for key, value in model.dict().items():
+            if isinstance(value, LiteModel):
+                # 如果是嵌套模型，先保存嵌套模型
+                nested_model_id = self.save(value)
+                # 保存嵌套模型的id, 以特殊前缀表示为模型
+                column_list.append(f'$id_{key}')
+                value_list.append(f'{value.__class__.__name__}_{nested_model_id}')
+            elif isinstance(value, list):
+                # 如果是列表，先保存列表中的所有嵌套模型，有可能有多种类型的嵌套模型
+                # 列表内存'ModelType_ModelId'，以特殊前缀表示为模型类型
+                nested_model_ids = []
+                for nested_model in value:
+                    nested_model_id = self.save(nested_model)
+                    nested_model_ids.append(f'{nested_model.__class__.__name__}_{nested_model_id}')
+                column_list.append(f'$ids_{key}')
+                value_list.append(nested_model_ids)
+
+        columns = ', '.join(column_list)
+        placeholders = ', '.join(['?' for _ in value_list])
+        values = tuple(value_list)
+        print(model.dict())
+        print(table_name, columns, placeholders, values)
+
+        # 插入数据
+        self.db.execute(f'INSERT INTO {table_name} ({columns}) VALUES ({placeholders})', values)
+        self.db.commit()
+        return self.db.execute(f'SELECT last_insert_rowid()').fetchone()[0]
+
+    def where(self, model_type: type(LiteModel) | str, conditions: str, *args, objectify: bool = True) -> list[LiteModel]:
+        """查询数据
+        Args:
+            objectify: bool: 是否将查询结果转换为模型
+            model_type: BaseModel
+            conditions: str
+            *args:
+            Returns:
+        """
+        table_name = self.get_model_table_name(model_type)
+        self.db.execute(f'CREATE TABLE IF NOT EXISTS {table_name} (id INTEGER PRIMARY KEY AUTOINCREMENT)')
+        return [self._convert_to_model(model_type, item) for item in self.db.execute(f'SELECT * FROM {table_name} WHERE {conditions}', args).fetchall()]
+
+    def first(self, model_type: type(LiteModel) | str, conditions: str, *args, objectify: bool = True):
+        """查询第一条数据"""
+        table_name = self.get_model_table_name(model_type)
+        self.db.execute(f'CREATE TABLE IF NOT EXISTS {table_name} (id INTEGER PRIMARY KEY AUTOINCREMENT)')
+        return self._convert_to_model(model_type, self.db.execute(f'SELECT * FROM {table_name} WHERE {conditions}', args).fetchone())
+
+    def all(self, model_type: type(LiteModel) | str, objectify: bool = True):
+        """查询所有数据"""
+        table_name = self.get_model_table_name(model_type)
+        self.db.execute(f'CREATE TABLE IF NOT EXISTS {table_name} (id INTEGER PRIMARY KEY AUTOINCREMENT)')
+        return [self._convert_to_model(model_type, item) for item in self.db.execute(f'SELECT * FROM {table_name}').fetchall()]
+
+    def update(self, model_type: type(LiteModel) | str, operation: str, conditions: str, *args):
+        """更新数据
+        Args:
+            model_type: BaseModel
+            operation: str: 更新操作
+            conditions: str: 查询条件
+            *args:
+        Returns:
+        """
+        table_name = self.get_model_table_name(model_type)
+        self.db.execute(f'CREATE TABLE IF NOT EXISTS {table_name} (id INTEGER PRIMARY KEY AUTOINCREMENT)')
+        self.db.execute(f'UPDATE {table_name} SET {operation} WHERE {conditions}', args)
+        self.db.commit()
+
+    def _convert_to_model(self, model_type: type(LiteModel), item: tuple) -> LiteModel:
+        """将查询结果转换为模型，处理嵌套模型"""
+        # 获取表中已有的字段，再用字段值构建字典
+        table_name = self.get_model_table_name(model_type)
+        cursor = self.db.execute(f"PRAGMA table_info({table_name})")
+        columns = [column_info[1] for column_info in cursor.fetchall()]
+        item_dict = dict(zip(columns, item))
+        # 遍历字典，处理嵌套模型
+        new_item_dict = {}
+        for key, value in item_dict.items():
+            if key.startswith('$id_'):
+                # 处理单个嵌套模型类型时从model_type中获取键
+                new_item_dict[key.replace('$id_', '')] = self.first(model_type.__annotations__[key.replace('$id_', '')], 'id = ?', value.split('_')[-1])
+            elif key.startswith('$ids_'):
+                # 处理多个嵌套模型类型使用eval获取数据库对应索引的键
+                new_item_dict[key.replace('$ids_', '')] = [self.first(eval(type_id.split('_')[0]), 'id = ?', type_id.split('_')[-1]) for type_id in value]
+            else:
+                new_item_dict[key] = value
+
+        return model_type(**new_item_dict)
diff --git a/src/liteyuki.py b/src/liteyuki.py
index 783ba76..f932cb7 100644
--- a/src/liteyuki.py
+++ b/src/liteyuki.py
@@ -1,9 +1,9 @@
 from typing import Any, Optional
-from nonebot import DOTENV_TYPE
 import nonebot
-from nonebot import logger
+from nonebot import DOTENV_TYPE
 from nonebot.adapters.onebot.v11 import Adapter as OnebotV11Adapter
 from nonebot.adapters.onebot.v12 import Adapter as OnebotV12Adapter
+from sqlalchemy import create_engine
 
 from src.api.utils import load_config
 
@@ -36,9 +36,9 @@ $$$$$$$$/ $$$$$$/    $$/    $$$$$$$$/     $$/      $$$$$$/  $$/   $$/ $$$$$$/ ''
     def run(self, *args, **kwargs):
         for adapter in adapters:
             self.driver.register_adapter(adapter)
-        self.nonebot.load_plugin('src.liteyuki_main')
-        self.nonebot.load_plugins('src/builtin')
-        self.nonebot.load_plugins('plugins')
+        self.nonebot.load_plugin('src.liteyuki_main')  # Load main plugin
+        self.nonebot.load_plugins('src/builtin')  # Load builtin plugins
+        self.nonebot.load_plugins('plugins')  # Load custom plugins
         # Todo: load from database
         self.nonebot.run(*args, **kwargs)