diff --git a/data_platform/strategy_base.py b/data_platform/strategy_base.py
new file mode 100644
index 000000000..b97b8e691
--- /dev/null
+++ b/data_platform/strategy_base.py
@@ -0,0 +1,57 @@
+"""
+BaseStrategy - 策略基类
+
+策略开发者只需继承此类并实现 generate_signals() 方法。
+回测引擎通过统一接口调用策略。
+
+Usage:
+    from data_platform.strategy_base import BaseStrategy
+
+    class MAStrategy(BaseStrategy):
+        def generate_signals(self, data: pd.DataFrame) -> pd.DataFrame:
+            data = data.copy()
+            data["ma5"] = data["close"].rolling(5).mean()
+            data["ma20"] = data["close"].rolling(20).mean()
+            data["signal"] = 0
+            data.loc[data["ma5"] > data["ma20"], "signal"] = 1   # 买入
+            data.loc[data["ma5"] < data["ma20"], "signal"] = -1  # 卖出
+            return data
+"""
+
+import pandas as pd
+from abc import ABC, abstractmethod
+
+
+class BaseStrategy(ABC):
+    """
+    策略基类 —— 所有回测策略必须继承此类
+
+    子类只需实现 generate_signals(data) -> data_with_signals
+    signal 列约定：
+        1  = 买入信号
+        -1 = 卖出信号
+        0  = 无操作
+    """
+
+    @abstractmethod
+    def generate_signals(self, data: pd.DataFrame) -> pd.DataFrame:
+        """
+        根据行情数据生成交易信号
+
+        Args:
+            data: 日线行情 DataFrame，至少包含 date, open, high, low, close, volume
+
+        Returns:
+            原始数据追加 signal 列（1=买, -1=卖, 0=无操作）
+        """
+        ...
+
+    @property
+    def name(self) -> str:
+        """策略名称，默认取类名"""
+        return self.__class__.__name__
+
+    @property
+    def description(self) -> str:
+        """策略描述，子类可覆盖"""
+        return ""