feat: Agent 运行时、对话 API、作业助手与引擎修复及前端执行超时

- agent_runtime 模块与 agent_chat API，前端 AgentChat 视图与路由对接
- workflow_engine: code 节点命名空间与 json 引用修复
- llm_service: 工具调用 extra_body（如 DeepSeek）
- create_homework_manager_agent / _3 脚本与测试脚本扩展
- frontend: WORKFLOW_EXECUTION_HTTP_TIMEOUT_MS、AgentChatPreview/MainLayout 等
- 文档：架构说明与自主 Agent 改造完成情况

Made-with: Cursor

backend/app/agent_runtime/schemas.py

@@ -0,0 +1,64 @@
+"""
+Agent Runtime 配置与数据结构 Schema
+"""
+from __future__ import annotations
+
+from typing import Any, Dict, List, Optional
+from pydantic import BaseModel, Field
+
+
+class AgentToolConfig(BaseModel):
+    """Agent 可用工具配置"""
+    # 若为空列表则使用全部已注册工具
+    include_tools: List[str] = Field(default_factory=list, description="允许的工具名称白名单")
+    exclude_tools: List[str] = Field(default_factory=list, description="排除的工具名称黑名单")
+
+
+class AgentMemoryConfig(BaseModel):
+    """Agent 记忆配置"""
+    enabled: bool = True
+    max_history_messages: int = 20  # 注入 LLM 的上文最大消息数
+    session_key: Optional[str] = None  # 会话标识，默认自动生成
+    persist_to_db: bool = True       # 是否写入 MySQL 长期记忆
+
+
+class AgentLLMConfig(BaseModel):
+    """Agent 模型配置"""
+    provider: str = "openai"        # openai / deepseek
+    model: str = "gpt-4o-mini"
+    temperature: float = 0.7
+    max_tokens: Optional[int] = None
+    api_key: Optional[str] = None
+    base_url: Optional[str] = None
+    max_iterations: int = 10        # ReAct 循环最大步数
+    request_timeout: float = 120.0
+    extra_body: Optional[Dict[str, Any]] = None
+
+
+class AgentConfig(BaseModel):
+    """Agent 完整配置"""
+    name: str = "default_agent"
+    system_prompt: str = "你是一个有用的AI助手。请使用可用工具来帮助用户完成任务。"
+    llm: AgentLLMConfig = Field(default_factory=AgentLLMConfig)
+    tools: AgentToolConfig = Field(default_factory=AgentToolConfig)
+    memory: AgentMemoryConfig = Field(default_factory=AgentMemoryConfig)
+    user_id: Optional[str] = None
+
+
+class AgentMessage(BaseModel):
+    """Agent 对话消息"""
+    role: str  # user / assistant / tool
+    content: str
+    tool_calls: Optional[List[Dict[str, Any]]] = None
+    tool_call_id: Optional[str] = None
+    name: Optional[str] = None
+
+
+class AgentResult(BaseModel):
+    """Agent 执行结果"""
+    success: bool = True
+    content: str = ""
+    truncated: bool = False
+    iterations_used: int = 0
+    tool_calls_made: int = 0
+    error: Optional[str] = None