从零开始构建 AI 智能体与多智能体系统：Python 实战完全指南

print("Welcome to my new project on AI Agents")

API_KEY=你的密钥粘贴在这里

pip install openaipip install python-dotenv

# 导入所需库from openai import OpenAIfrom dotenv import load_dotenvimport os
# 加载 .env 文件中的环境变量load_dotenv()api_key = os.getenv("API_KEY")
# 创建 AI 客户端——这是你的程序与 AI 模型之间的"信使"client = OpenAI(    api_key=api_key,    base_url="https://openrouter.ai/api/v1")
# 无限循环，让智能体持续运行while True:    # 接收用户输入    question = input("You: ")    print("Thinking...\n")
    # 发送请求到 AI 模型    response = client.chat.completions.create(        model="baidu/cobuddy:free",  # 选择的 AI 模型        messages=[            {                "role": "system",                "content": "You are a helpful educational tutor."            },            {                "role": "user",                "content": question            }        ]    )
    # 提取并打印 AI 的回答    answer = response.choices[0].message.content    print("\nAI:", answer)    print("\n-------------------\n")

用户输入旅行需求    ↓Research Agent → 调研目的地信息    ↓Activity Agent → 制定活动计划    ↓Budget Agent → 计算预算    ↓Final Agent → 生成完整旅行方案    ↓输出给用户

from openai import OpenAIfrom dotenv import load_dotenvimport os
# 加载环境变量load_dotenv()api_key = os.getenv("API_KEY")
# 创建 AI 客户端，连接到 OpenRouterclient = OpenAI(    base_url="https://openrouter.ai/api/v1",    api_key=api_key)

classAgent:    def __init__(self, name, role):        self.name = name      # 智能体的名字        self.role = role      # 智能体的角色（告诉 AI 它应该怎么表现）
    def run(self, task):        """让智能体执行任务的核心方法"""        print(f"{self.name} is working...")
        response = client.chat.completions.create(            model="gpt-4.1-mini",  # 使用 GPT-4.1-mini 模型            messages=[                {                    "role": "system",                    "content": self.role  # 系统提示：定义智能体的行为                },                {                    "role": "user",                    "content": task       # 用户输入的任务                }            ],            max_tokens=1200  # 限制返回的最大 token 数        )
        # 提取 AI 生成的回答        return response.choices[0].message.content

# 🔬 调研智能体research_agent = Agent(    "Research Agent",    """You are an expert travel researcher. Your job:    - Find popular attractions    - Find hidden gems    - Suggest local experiences    - Recommend best places""")
# 📋 活动规划智能体activity_agent = Agent(    "Activity Planner Agent",    """You are a professional travel planner. Your job:    - Create daily activities    - Plan sightseeing    - Recommend food experiences    - Organize activities logically""")
# 💰 预算智能体budget_agent = Agent(    "Budget Agent",    """You are a travel budget expert. Calculate:    - Estimated flight cost    - Visa requirements    - Visa fees if needed    - Hotel cost    - Food expenses    - Transport cost    - Activity costs    Create an approximate total trip budget. Keep response short.""")
# 🎯 最终助手final_agent = Agent(    "Final Travel Assistant",    """You are a professional travel planner. Create the final itinerary. Include:    1. Short trip overview    2. Visa information    3. Estimated flight cost    4. Day-wise plan    5. Food suggestions    6. Total estimated budget    Keep everything under 700 words.""")

starting_location = input("Where are you flying from? ")destination = input("Where do you want to travel? ")days = input("How many days is your trip? ")travelers = input("How many travelers? ")budget = input("What is your budget? (low/medium/high) ")interests = input("What are your interests? ")
# 将所有信息整合成一个完整的任务描述user_request = f"""Create a travel plan with these details:- Flying From: {starting_location}- Destination: {destination}- Trip Duration: {days} days- Number of Travelers: {travelers}- Budget Level: {budget}- Interests: {interests}
Include:- Visa requirements- Estimated flight cost- Places to visit- Activities- Food recommendations- Total estimated budget"""
print("\nCreating your AI travel plan...\n")

# 第1步：调研智能体工作research = research_agent.run(user_request)print("\n--- Research Completed ---")
# 第2步：活动规划智能体基于调研结果工作activities = activity_agent.run(research)print("\n--- Activities Planned ---")
# 第3步：预算智能体基于活动计划计算费用budget = budget_agent.run(activities)print("\n--- Budget Created ---")
# 第4步：最终助手整合所有信息，生成完整行程final_plan = final_agent.run(f"""Research: {research}Activities: {activities}Budget: {budget}Create final travel plan.""")
print("\n==========================")print(" FINAL TRAVEL PLAN")print("==========================\n")print(final_plan)

user_request（包含所有旅行信息）    ↓ 传递给Research Agent → 输出："伊斯坦布尔有圣索菲亚大教堂、托普卡帕宫、              伊斯坦布尔水族馆、KidZania 等景点，建议3天游览..."    ↓ 这个输出作为输入传递给Activity Agent → 输出："第1天上午参观圣索菲亚大教堂和托普卡帕宫，              下午在苏丹艾哈迈德广场休息，晚上坐博斯普鲁斯游船晚餐..."    ↓ 这个活动计划作为输入传递给Budget Agent → 输出："机票约$1400-$1800（4人），签证$200，住宿$600-$800，              餐饮$300-$400，交通$100，活动$300，总计约$3100-$3800"    ↓ 调研+活动+预算 三者合并后传递给


    
Final Agent → 输出：一份完整的3天家庭友好型伊斯坦布尔旅行行程

Where are you flying from? islamabadWhere do you want to travel? istanbulHow many days is your trip? 3How many travelers? 4What is your budget? (low/medium/high) $4kWhat are your interests? kid friendly
Creating your AI travel plan...
Research Agent is working...--- Research Completed ---Activity Planner Agent is working...--- Activities Planned ---Budget Agent is working...--- Budget Created ---Final Travel Assistant is working...
========================== FINAL TRAVEL PLAN==========================
### 3-Day Family-Friendly Itinerary: Islamabad to Istanbul
#### Trip OverviewDiscover Istanbul's captivating blend of history, culture, and family-friendly fun...
#### Visa Information- For Pakistani Citizens: Turkish e-Visa required- Cost: Approx. $50 per person
#### Estimated Flight Cost- $350 - $450 per person in economy class- For 4 travelers: Approx. $1,400 - $1,800 total
#### Day-wise ItineraryDay 1: Historic and Iconic Sights  - Morning: Hagia Sophia, Topkapi Palace  - Afternoon: Sultanahmet Square, tram ride  - Evening: Bosphorus dinner cruise
Day 2: Interactive Fun and Exploration  - Morning: Istanbul Aquarium  - Afternoon: KidZania Istanbul  - Evening


    
: Turkish ice cream show
Day 3: Parks, Museums, Markets  - Morning: Gülhane Park, Rahmi M. Koç Museum  - Afternoon: Grand Bazaar, street food  - Evening: Dinner at Çiya Sofrası
#### Estimated Budget (4 Persons)| Category | Estimated Cost (USD) ||----------|---------------------|| Flights | $1,400 - $1,800 || Visa Fees | $200 || Accommodation | $600 - $800 || Food | $300 - $400 || Transport | $100 || Activities | $300 || **Total** | **~$3,100 - $3,800** |