refactor : FastAPI 懒加载单例，消除模块级副作用

- import astrai.inference 不再在模块加载时创建 FastAPI 实例
- 路由移至 APIRouter；get_app() 首次调用时懒构造单例
- _create_engine 和 run_server 的 param_path 改为必填
- 更新测试改用 get_app() 替代模块级 app

astrai/inference/api/__init__.py

@@ -1,4 +1,8 @@
-"""Inference API: protocol handler, stop checker, and FastAPI server."""
+"""Inference API: protocol handler, stop checker, and FastAPI server.
+
+``app`` is no longer a module-level global. Use :func:`get_app` to access the
+lazy singleton FastAPI instance.
+"""
 
 from astrai.inference.api.protocol import GenContext, ProtocolHandler, StopChecker
 from astrai.inference.api.server import (
@@ -6,7 +10,7 @@ from astrai.inference.api.server import (
     ChatCompletionRequest,
     ChatMessage,
     MessagesRequest,
-    app,
+    get_app,
     run_server,
 )
 
@@ -18,6 +22,6 @@ __all__ = [
     "ChatCompletionRequest",
     "ChatMessage",
     "MessagesRequest",
-    "app",
+    "get_app",
     "run_server",
 ]