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/api/agent_chat.py

@@ -0,0 +1,137 @@
+"""
+Agent 独立聊天 API — 不依赖工作流 DAG，直接与 Agent Runtime 对话。
+
+POST /api/v1/agent-chat/bare
+    {"message": "你好，帮我..."}
+    → {"content": "...", "iterations": 3, "tool_calls": 5}
+"""
+from __future__ import annotations
+
+import logging
+from typing import Any, Dict, Optional
+from fastapi import APIRouter, Depends, HTTPException
+from pydantic import BaseModel
+
+from app.core.database import get_db
+from sqlalchemy.orm import Session
+from app.api.auth import get_current_user
+from app.models.user import User
+from app.models.agent import Agent
+from app.agent_runtime import (
+    AgentRuntime,
+    AgentConfig,
+    AgentLLMConfig,
+    AgentToolConfig,
+)
+from app.core.config import settings
+
+logger = logging.getLogger(__name__)
+router = APIRouter(prefix="/api/v1/agent-chat", tags=["agent-chat"])
+
+
+class ChatRequest(BaseModel):
+    message: str
+    session_id: Optional[str] = None
+    model: Optional[str] = None
+    temperature: Optional[float] = None
+    max_iterations: Optional[int] = None
+
+
+class ChatResponse(BaseModel):
+    content: str
+    iterations_used: int
+    tool_calls_made: int
+    truncated: bool
+    session_id: str
+    agent_id: Optional[str] = None
+
+
+@router.post("/bare", response_model=ChatResponse)
+async def chat_bare(
+    req: ChatRequest,
+    current_user: User = Depends(get_current_user),
+):
+    """无需 Agent 配置，使用默认设置直接对话。"""
+    config = AgentConfig(
+        name="bare_agent",
+        system_prompt="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(
+            model=req.model or (
+                "gpt-4o-mini" if settings.OPENAI_API_KEY and settings.OPENAI_API_KEY != "your-openai-api-key"
+                else "deepseek-v4-flash"
+            ),
+            temperature=req.temperature or 0.7,
+            max_iterations=req.max_iterations or 10,
+        ),
+        user_id=current_user.id,
+    )
+    runtime = AgentRuntime(config=config)
+    result = await runtime.run(req.message)
+
+    return ChatResponse(
+        content=result.content,
+        iterations_used=result.iterations_used,
+        tool_calls_made=result.tool_calls_made,
+        truncated=result.truncated,
+        session_id=runtime.context.session_id,
+    )
+
+
+@router.post("/{agent_id}", response_model=ChatResponse)
+async def chat_with_agent(
+    agent_id: str,
+    req: ChatRequest,
+    current_user: User = Depends(get_current_user),
+    db: Session = Depends(get_db),
+):
+    """与指定的 Agent 对话。Agent 的工作流配置会用于构建 Runtime。"""
+    agent = db.query(Agent).filter(Agent.id == agent_id).first()
+    if not agent:
+        raise HTTPException(status_code=404, detail="Agent 不存在")
+    if agent.user_id and agent.user_id != current_user.id and current_user.role != "admin":
+        raise HTTPException(status_code=403, detail="无权访问该 Agent")
+
+    # 从 Agent 配置构建 Runtime
+    wc = agent.workflow_config or {}
+    nodes = wc.get("nodes", [])
+    # 查找 agent 节点的配置（或第一个 llm 节点的配置）
+    agent_node_cfg = _find_agent_node_config(nodes)
+
+    config = AgentConfig(
+        name=agent.name,
+        system_prompt=agent_node_cfg.get("system_prompt") or agent.description or "你是一个有用的AI助手。",
+        llm=AgentLLMConfig(
+            provider=agent_node_cfg.get("provider", "openai"),
+            model=req.model or agent_node_cfg.get("model", "gpt-4o-mini"),
+            temperature=req.temperature or float(agent_node_cfg.get("temperature", 0.7)),
+            max_iterations=req.max_iterations or int(agent_node_cfg.get("max_iterations", 10)),
+        ),
+        tools=AgentToolConfig(
+            include_tools=agent_node_cfg.get("tools", []),
+            exclude_tools=agent_node_cfg.get("exclude_tools", []),
+        ),
+        user_id=current_user.id,
+    )
+
+    runtime = AgentRuntime(config=config)
+    result = await runtime.run(req.message)
+
+    return ChatResponse(
+        content=result.content,
+        iterations_used=result.iterations_used,
+        tool_calls_made=result.tool_calls_made,
+        truncated=result.truncated,
+        session_id=runtime.context.session_id,
+        agent_id=agent_id,
+    )
+
+
+def _find_agent_node_config(nodes: list) -> Dict[str, Any]:
+    """从工作流节点列表中查找第一个 agent 类型或 llm 类型的节点配置。"""
+    if not nodes:
+        return {}
+    for node in nodes:
+        typ = node.get("type", "")
+        if typ in ("agent", "llm", "template"):
+            return node.get("data") or {}
+    return {}