2450550235-debug/wechat-auto-send/SKILL.md
#!/usr/bin/env python # -*- coding: utf-8 -*- """ 微信自动发送消息 - 命令行参数版 技能版本:v1.0 创建时间:2026-03-18 """ import pyautogui import pygetwindow import pyperclip import time import sys import argparse def setup_encoding(): """尝试设置控制台编码""" try: import io sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding='utf-8') except: pass def check_dependencies(): """检查依赖""" try: import pyautogui import pygetwindow import pyperclip ret
npx skillsauth add openclaw/skills 2450550235-debug/wechat-auto-sendInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
#!/usr/bin/env python
""" 微信自动发送消息 - 命令行参数版 技能版本:v1.0 创建时间:2026-03-18 """
import pyautogui import pygetwindow import pyperclip import time import sys import argparse
def setup_encoding(): """尝试设置控制台编码""" try: import io sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding='utf-8') except: pass
def check_dependencies(): """检查依赖""" try: import pyautogui import pygetwindow import pyperclip return True except ImportError as e: print(f"[错误] 缺少依赖: {e}") print("请安装: pip install pyautogui pygetwindow pyperclip") return False
def focus_wechat(): """聚焦微信窗口""" try: wechat_windows = pygetwindow.getWindowsWithTitle('微信') if wechat_windows: window = wechat_windows[0] if window.isMinimized: window.restore() window.activate() time.sleep(1.5) return True else: print("[错误] 未找到微信窗口,请确保微信已打开并可见") return False except Exception as e: print(f"[警告] 聚焦窗口失败: {e}") return True # 继续尝试
def send_to_contact(contact_name, message, wait_time=1.5): """发送消息给指定联系人""" print(f"准备发送给: {contact_name}") print(f"消息内容: {message}")
# 1. 复制联系人
pyperclip.copy(contact_name)
print(" [1] 已复制联系人")
time.sleep(0.5)
# 2. 打开搜索框
pyautogui.hotkey('ctrl', 'f')
print(" [2] 按 Ctrl+F 打开搜索框")
time.sleep(1.0)
# 3. 粘贴联系人
pyautogui.hotkey('ctrl', 'v')
print(" [3] 按 Ctrl+V 粘贴联系人")
time.sleep(0.8)
# 4. 按 Enter 选择
pyautogui.press('enter')
print(" [4] 按 Enter 进入聊天")
time.sleep(wait_time)
# 5. 复制消息
pyperclip.copy(message)
print(" [5] 已复制消息")
time.sleep(0.5)
# 6. 粘贴消息
pyautogui.hotkey('ctrl', 'v')
print(" [6] 按 Ctrl+V 粘贴消息")
time.sleep(0.5)
# 7. 发送消息
pyautogui.press('enter')
print(" [7] 按 Enter 发送")
time.sleep(1.0)
return True
def main(): # 设置编码 setup_encoding()
# 解析命令行参数
parser = argparse.ArgumentParser(
description='微信自动发送消息工具',
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
示例: %(prog)s "文件传输助手" "这是一条测试消息" %(prog)s "张三" "会议改到下午3点" --wait 2 --countdown 3 """ )
parser.add_argument('contact', help='联系人名称(备注或昵称)')
parser.add_argument('message', help='要发送的消息内容')
parser.add_argument('--wait', type=float, default=1.5,
help='操作等待时间(秒,默认1.5)')
parser.add_argument('--countdown', type=int, default=5,
help='开始前倒计时秒数(默认5)')
parser.add_argument('--verbose', action='store_true',
help='显示详细执行信息')
args = parser.parse_args()
print("=" * 60)
print("微信自动发送消息工具")
print("=" * 60)
# 检查依赖
if not check_dependencies():
return 1
if args.verbose:
print(f"\n配置信息:")
print(f" 联系人: {args.contact}")
print(f" 消息: {args.message}")
print(f" 等待时间: {args.wait}秒")
print(f" 倒计时: {args.countdown}秒")
print("\n[准备提示]")
print("1. 请确保微信已经打开并处于前台")
print("2. 确保可以看到微信主界面")
print("3. 脚本将在倒计时后开始执行")
print("4. 执行过程中请不要操作鼠标键盘")
print("5. 中断方法: 快速移动鼠标到屏幕左上角")
# 倒计时
print(f"\n倒计时 {args.countdown} 秒开始...")
for i in range(args.countdown, 0, -1):
print(f" {i}...")
time.sleep(1)
print("\n开始执行...")
try:
# 聚焦微信
if not focus_wechat():
print("[错误] 无法聚焦微信窗口")
return 1
# 发送消息
success = send_to_contact(args.contact, args.message, args.wait)
if success:
print("\n" + "=" * 60)
print("[成功] 消息发送完成!")
print("=" * 60)
return 0
else:
print("\n[失败] 消息发送失败")
return 1
except KeyboardInterrupt:
print("\n[用户中断] 脚本被中断")
return 1
except Exception as e:
print(f"\n[错误] 执行出错: {e}")
return 1
finally:
pyautogui.FAILSAFE = True
if name == "main": # 安全设置 pyautogui.FAILSAFE = True # 鼠标移到左上角可中断 pyautogui.PAUSE = 0.1 # 每个动作后暂停0.1秒
sys.exit(main())
tools
Use when the user wants to connect to, test, or use the McDonalds service at mcp.mcd.cn, including checking authentication, probing MCP endpoints, listing tools, or calling McDonalds MCP tools through a reusable local CLI.
development
Web scraping platform — Twitter/X data, Vinted marketplace, and general web scraping API
development
SlowMist AI Agent Security Review — comprehensive security framework for skills, repositories, URLs, on-chain addresses, and products (Claude Code version)
data-ai
去除中文文本中的 AI 写作痕迹,使其读起来自然。基于维基百科 AI 写作特征指南,检测 24 种 AI 模式。触发词:humanizer-cn、去除 AI 痕迹、去除 AI 写作痕迹、中文文本人性化。