# 如何用ChatGPT写一个OKX抢开盘脚本

By [Daniel ](https://paragraph.com/@daniel-32) · 2023-03-08

---

本文所有代码全部都是chatgpt写的。 前两天留言问大家想要什么脚本，看到了有OKX开盘的，觉得这个不错

本文主要是教大家怎么描述需求让AI写代码～ 代码块上加粗部分都是我的对话语句，代码文件全部chatgpt写的。

**research this python sdk** [**https://github.com/coinrising/okex-api-v5**](https://github.com/coinrising/okex-api-v5) **Assist me in writing Python code that calls the OKX API and utilizes its SDK.**

    from okex.account_api import AccountAPI
    from okex.market_data_api import MarketDataAPI
    from okex.trade_api import TradeAPI
    from okex.const import *
    
    # Set up authentication credentials
    api_key = 'YOUR_API_KEY'
    secret_key = 'YOUR_SECRET_KEY'
    passphrase = 'YOUR_PASSPHRASE'
    
    # Create API clients
    account_client = AccountAPI(api_key, secret_key, passphrase, True)
    trade_client = TradeAPI(api_key, secret_key, passphrase, True)
    market_data_client = MarketDataAPI()
    
    # Get the latest market price of BTC
    response = market_data_client.get_ticker('BTC-USDT')
    latest_price = response['last']
    print(f"The latest price of BTC is {latest_price} USDT")
    

首先让他学习了okx的文档，这里已经识别出来怎么调用okex的Api来进行市场操作,但是我们的需求是抢开盘，那么告诉他。

**I want to monitor a token for its listing and buy it as soon as it is listed.**

    import schedule
    import time
    from okex.trade_api import TradeAPI
    from okex.account_api import AccountAPI
    from okex.const import *
    
    # Set up authentication credentials
    api_key = 'YOUR_API_KEY'
    secret_key = 'YOUR_SECRET_KEY'
    passphrase = 'YOUR_PASSPHRASE'
    
    # Set up trade parameters
    symbol = 'TOKEN-USDT'  # Replace with your desired token symbol
    amount = 100  # Replace with your desired order amount in USDT
    
    # Create API clients
    account_client = AccountAPI(api_key, secret_key, passphrase, True)
    trade_client = TradeAPI(api_key, secret_key, passphrase, True)
    
    def buy_token():
        try:
            # Check if the token is listed
            response = trade_client.get_specific_public_symbol(symbol)
            if response['state'] == 'live':
                # If the token is listed, place a market order to buy it
                response = trade_client.place_market_order(symbol, SIDE_BUY, funds=amount)
                print(response)
        except Exception as e:
            print(f"An error occurred: {e}")
    
    # Schedule the buy_token function to be called every minute
    schedule.every(1).minutes.do(buy_token)
    
    # Run the scheduler loop
    while True:
        schedule.run_pending()
        time.sleep(1)
    

这里看到已经在监控代币是否上线了，但是购买的时候间隔太长，并且一直循环购买，我想要在过了某一时间段之后疯狂请求购买，并且都变成参数可调。

**Modify to scan the token online at millisecond level after reaching a certain time, and purchase it immediately once it is available.Do not use schedule.every().day, directly start the program when judging that the time is greater than something.**

    import time
    import argparse
    from datetime import datetime
    from okex.trade_api import TradeAPI
    from okex.account_api import AccountAPI
    from okex.const import *
    
    # Set up authentication credentials
    api_key = 'YOUR_API_KEY'
    secret_key = 'YOUR_SECRET_KEY'
    passphrase = 'YOUR_PASSPHRASE'
    
    # Set up trade parameters
    symbol = 'TOKEN-USDT'  # Replace with your desired token symbol
    amount = 100  # Replace with your desired order amount in USDT
    
    # Create API clients
    account_client = AccountAPI(api_key, secret_key, passphrase, True)
    trade_client = TradeAPI(api_key, secret_key, passphrase, True)
    
    def buy_token():
        try:
            # Check if the token is listed
            response = trade_client.get_specific_public_symbol(symbol)
            if response['state'] == 'live':
                # If the token is listed, place a market order to buy it
                response = trade_client.place_market_order(symbol, SIDE_BUY, funds=amount)
                print(response)
                return True
        except Exception as e:
            print(f"An error occurred: {e}")
            return False
    
    if __name__ == '__main__':
        # Parse command-line arguments
        parser = argparse.ArgumentParser(description='Monitor token listing and buy immediately')
        parser.add_argument('year', type=int, help='Listing year')
        parser.add_argument('month', type=int, help='Listing month')
        parser.add_argument('day', type=int, help='Listing day')
        parser.add_argument('hour', type=int, help='Listing hour')
        parser.add_argument('minute', type=int, help='Listing minute')
        parser.add_argument('second', type=int, help='Listing second')
        args = parser.parse_args()
    
        # Convert command-line arguments to datetime object
        listing_time = datetime(args.year, args.month, args.day, args.hour, args.minute, args.second)
    
        # Monitor the token listing
        while datetime.now() < listing_time:
            time.sleep(1)
    
        while not buy_token():
            time.sleep(0.1)
    

这里已经基本上可用了，能监控交易对，能配置什么时候开始，（咱们只需要配置在开盘前3分钟就疯狂监控就可以了）但是我想让他更快一点呢？多线程疯狂请求。

**Can we enable multi-threading during purchase to make it faster?**

    import time
    import argparse
    from datetime import datetime
    from threading import Thread
    from okex.trade_api import TradeAPI
    from okex.account_api import AccountAPI
    from okex.const import *
    
    # Set up authentication credentials
    api_key = 'YOUR_API_KEY'
    secret_key = 'YOUR_SECRET_KEY'
    passphrase = 'YOUR_PASSPHRASE'
    
    # Set up trade parameters
    symbol = 'TOKEN-USDT'  # Replace with your desired token symbol
    amount = 100  # Replace with your desired order amount in USDT
    num_threads = 5  # Replace with the number of threads to use
    
    # Create API clients
    account_client = AccountAPI(api_key, secret_key, passphrase, True)
    trade_client = TradeAPI(api_key, secret_key, passphrase, True)
    
    def buy_token():
        try:
            # Check if the token is listed
            response = trade_client.get_specific_public_symbol(symbol)
            if response['state'] == 'live':
                # If the token is listed, place a market order to buy it
                response = trade_client.place_market_order(symbol, SIDE_BUY, funds=amount)
                print(response)
                return True
        except Exception as e:
            print(f"An error occurred: {e}")
            return False
    
    if __name__ == '__main__':
        # Parse command-line arguments
        parser = argparse.ArgumentParser(description='Monitor token listing and buy immediately')
        parser.add_argument('year', type=int, help='Listing year')
        parser.add_argument('month', type=int, help='Listing month')
        parser.add_argument('day', type=int, help='Listing day')
        parser.add_argument('hour', type=int, help='Listing hour')
        parser.add_argument('minute', type=int, help='Listing minute')
        parser.add_argument('second', type=int, help='Listing second')
        args = parser.parse_args()
    
        # Convert command-line arguments to datetime object
        listing_time = datetime(args.year, args.month, args.day, args.hour, args.minute, args.second)
    
        # Wait until the listing time
        while datetime.now() < listing_time:
            time.sleep(1)
    
        # Buy the token using multiple threads
        threads = []
        for i in range(num_threads):
            thread = Thread(target=buy_token)
            thread.start()
            threads.append(thread)
    
        # Wait for all threads to complete
        for thread in threads:
            thread.join()
    

到这里基本上逻辑看起来完全可以用了，只需要稍微修改一下配置。 问题来了，是不是真的可用呢？ 答案是否的，虽然逻辑都看起来对了，但是这个是调用okex的sdk，okex并没有官方sdk，而都是社区自己写的，这个时候就需要去找sdk的api是不是跟代码里面调用的一样的呢，不一样的话就需要修改成一样的，这个就需要稍微懂技术的人来修改了。

**综上可得，Chatgpt确实可以写代码，并且写的逻辑完全没问题，在写模块化组件，CRUD，常规的工具类方面很强，但是触及到各种第三方调用，整体架构之类的还需要程序员自己修改。当然也可以直接给他模版，让他学习。**

---

*Originally published on [Daniel ](https://paragraph.com/@daniel-32/chatgpt-okx)*