style: 统一代码格式和字符串引号使用

refactor: 优化代码结构和可读性

feat: 添加http_limit模块实现请求限制功能

fix: 修复异步任务中使用time.sleep的问题

chore: 更新依赖项并添加pytest测试框架

docs: 更新项目描述信息

perf: 优化Redis序列化方式使用JSON替代pickle

test: 添加测试相关配置和依赖

backend/app/plugin/module_application/ai/tools/agent_demo.py

@@ -17,7 +17,12 @@ from langchain.agents.structured_output import (
     ToolStrategy,
 )
 from langchain.chat_models import init_chat_model
-from langchain.messages import AIMessage, HumanMessage, RemoveMessage, SystemMessage
+from langchain.messages import (
+    AIMessage,
+    HumanMessage,
+    RemoveMessage,
+    SystemMessage,
+)
 from langchain.tools import ToolRuntime, tool
 from langgraph.checkpoint.memory import InMemorySaver
 from langgraph.graph.message import REMOVE_ALL_MESSAGES
@@ -46,6 +51,7 @@ def get_weather_for_location(city: str) -> str:
 @dataclass
 class Context:
     """Custom runtime context schema."""
+
     user_id: str
 
 
@@ -57,17 +63,13 @@ def get_user_location(runtime: ToolRuntime[Context]) -> str:
 
 
 # =================定义模型=================
-model = init_chat_model(
-    "claude-sonnet-4-5-20250929",
-    temperature=0.5,
-    timeout=10,
-    max_tokens=1000
-)
+model = init_chat_model("claude-sonnet-4-5-20250929", temperature=0.5, timeout=10, max_tokens=1000)
 
 
 # =================定义响应模型=================
 class ResponseFormat(BaseModel):
     """Response schema for the agent."""
+
     rating: int | None = Field(description="Rating from 1-5", ge=1, le=5)
     comment: str = Field(description="Review comment")
     punny_response: str
@@ -98,28 +100,21 @@ def trim_messages(state: AgentState, runtime: Runtime) -> dict[str, Any] | None:
     recent_messages = messages[-3:] if len(messages) % 2 == 0 else messages[-4:]
     new_messages = [first_msg, *recent_messages]
 
-    return {
-        "messages": [
-            RemoveMessage(id=REMOVE_ALL_MESSAGES),
-            *new_messages
-        ]
-    }
+    return {"messages": [RemoveMessage(id=REMOVE_ALL_MESSAGES), *new_messages]}
 
 
 @after_model
 def validate_response(state: AgentState, runtime: Runtime) -> dict | None:
     """Remove messages containing sensitive words."""
-    STOP_WORDS = ["password", "secret"]
     last_message = state["messages"][-1]
-    if any(word in last_message.content for word in STOP_WORDS):
+    if any(word in last_message.content for word in ["password", "secret"]):
         return {"messages": [RemoveMessage(id=last_message.id or "")]}
     return None
 
 
 @wrap_model_call
 def inject_file_context(
-    request: ModelRequest,
-    handler: Callable[[ModelRequest], ModelResponse]
+    request: ModelRequest, handler: Callable[[ModelRequest], ModelResponse]
 ) -> ModelResponse:
     """Inject context about files user has uploaded this session."""
     # Read from State: get uploaded files metadata
@@ -127,7 +122,9 @@ def inject_file_context(
 
     if uploaded_files:
         # Build context about available files
-        file_descriptions = [f"- {file['name']} ({file['type']}): {file['summary']}" for file in uploaded_files]
+        file_descriptions = [
+            f"- {file['name']} ({file['type']}): {file['summary']}" for file in uploaded_files
+        ]
 
         file_context = f"""Files you have access to in this conversation:
 {chr(10).join(file_descriptions)}
@@ -155,10 +152,18 @@ agent = create_agent(
     model=model,
     system_prompt=SYSTEM_PROMPT,
     tools=[get_user_location, get_weather_for_location],
-    middleware=[dynamic_system_prompt, trim_messages, validate_response, inject_file_context],
+    middleware=[
+        dynamic_system_prompt,
+        trim_messages,
+        validate_response,
+        inject_file_context,
+    ],
     context_schema=Context,
-    response_format=ToolStrategy(schema=ResponseFormat, handle_errors=(ValueError, TypeError, custom_error_handler)),
-    checkpointer=checkpointer
+    response_format=ToolStrategy(
+        schema=ResponseFormat,
+        handle_errors=(ValueError, TypeError, custom_error_handler),
+    ),
+    checkpointer=checkpointer,
 )
 
 # =================定义线程=================
@@ -167,18 +172,14 @@ config = {"configurable": {"thread_id": "1"}}
 messages = [
     SystemMessage("You are a poetry expert"),
     HumanMessage("Write a haiku about spring"),
-    AIMessage("Cherry blossoms bloom...")
+    AIMessage("Cherry blossoms bloom..."),
 ]
 
 # =================运行智能体=================
-response = agent.invoke(
-    input=messages,
-    config=config,
-    context=Context(user_id="1")
-)
+response = agent.invoke(input=messages, config=config, context=Context(user_id="1"))
 
 # =================解析响应=================
-print(response['structured_response'])
+print(response["structured_response"])
 # ResponseFormat(
 #     punny_response="Florida is still having a 'sun-derful' day! The sunshine is playing 'ray-dio' hits all day long! I'd say it's the perfect weather for some 'solar-bration'! If you were hoping for rain, I'm afraid that idea is all 'washed up' - the forecast remains 'clear-ly' brilliant!",
 #     weather_conditions="It's always sunny in Florida!"

backend/app/plugin/module_application/ai/tools/ai_util.py

@@ -5,6 +5,7 @@ from langchain_core.messages import HumanMessage, SystemMessage
 from langchain_openai import ChatOpenAI
 
 from app.config.setting import settings
+from app.core.exceptions import CustomException
 from app.core.logger import log
 
 
@@ -16,14 +17,14 @@ class AIClient:
     def __init__(self) -> None:
         # 使用LangChain的ChatOpenAI类
         self.model = ChatOpenAI(
-            api_key=settings.OPENAI_API_KEY,
+            api_key=lambda: settings.OPENAI_API_KEY,
             model=settings.OPENAI_MODEL,
             base_url=settings.OPENAI_BASE_URL,
             temperature=0.7,
-            streaming=True
+            streaming=True,
         )
 
-    async def process(self, query: str)  -> AsyncGenerator[str, Any]:
+    async def process(self, query: str) -> AsyncGenerator[str, Any]:
         """
         处理查询并返回流式响应
 
@@ -33,13 +34,15 @@ class AIClient:
         返回:
         - AsyncGenerator[str, Any]: 流式响应内容。
         """
-        system_prompt = """你是一个有用的AI助手，可以帮助用户回答问题和提供帮助。请用中文回答用户的问题。"""
+        system_prompt = (
+            """你是一个有用的AI助手，可以帮助用户回答问题和提供帮助。请用中文回答用户的问题。"""
+        )
 
         try:
             # 使用LangChain的异步流式生成
             messages = [
                 SystemMessage(content=system_prompt),
-                HumanMessage(content=query)
+                HumanMessage(content=query),
             ]
 
             # 使用LangChain的流式响应
@@ -66,24 +69,33 @@ class AIClient:
                 error_code = err.get("code")
                 message = err.get("message")
         except Exception:
-            # 忽略解析失败
-            pass
+            raise CustomException(f"解析 OpenAI 错误失败: {e!s}")
 
         text = str(e)
         msg = message or text
 
         # 特定错误映射
         # 欠费/账户状态异常
-        if (error_code == "Arrearage") or (error_type == "Arrearage") or ("in good standing" in (msg or "")):
+        if (
+            (error_code == "Arrearage")
+            or (error_type == "Arrearage")
+            or ("in good standing" in (msg or ""))
+        ):
             return "账户欠费或结算异常，访问被拒绝。请检查账号状态或更换有效的 API Key。"
         # 鉴权失败
         if status_code == 401 or "invalid api key" in msg.lower():
             return "鉴权失败，API Key 无效或已过期。请检查系统配置中的 API Key。"
         # 权限不足或被拒绝
-        if status_code == 403 or error_type in {"PermissionDenied", "permission_denied"}:
+        if status_code == 403 or error_type in {
+            "PermissionDenied",
+            "permission_denied",
+        }:
             return "访问被拒绝，权限不足或账号受限。请检查账户权限设置。"
         # 配额不足或限流
-        if status_code == 429 or error_type in {"insufficient_quota", "rate_limit_exceeded"}:
+        if status_code == 429 or error_type in {
+            "insufficient_quota",
+            "rate_limit_exceeded",
+        }:
             return "请求过于频繁或配额已用尽。请稍后重试或提升账户配额。"
         # 客户端错误
         if status_code == 400: