wechat-qwen-reply
WeChat chat reader + auto-reply (Qwen-VL vision + AHK send). Supports fast/slow capture, group nickname labels, file/red-packet cards, and filtering system messages on Windows.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/chenxundaozu/wechat-qwen-replyWhat This Skill Does
The wechat-qwen-reply skill is an advanced automation bridge for Windows users that enables an AI agent to interact with the WeChat PC client. By combining computer vision capabilities (via Qwen-VL) with local scripting (PowerShell and AutoHotkey), it allows the agent to "see" the current conversation, interpret messages, and draft automated replies. The skill is designed to handle the nuances of chat interfaces, such as distinguishing between sender bubbles, identifying system notifications, and parsing card-based content like files or red packets, all while bypassing the manual constraints of the WeChat UI.
Installation
To install this skill, use the OpenClaw CLI:
clawhub install openclaw/skills/skills/chenxundaozu/wechat-qwen-reply
Ensure you have the following prerequisites installed on your system:
- Python 3.12
- AHK (AutoHotkey) v2
- A valid DashScope API Key saved at
C:\Users\chenxun\.nanobot\workspace\.secrets\dashscope_api_key.txt
Use Cases
- Automated Customer Service: Monitor a specific WeChat group or contact to provide instant, AI-generated responses based on historical context.
- Information Extraction: Automatically extract and summarize unstructured chat data from group discussions into local text files for easy reference.
- Notifications Aggregation: Keep track of incoming file transmissions or urgent messages in professional groups without constantly checking the monitor.
- Intelligent Response Handling: Automatically acknowledge file transfers or red packets with polite, predefined messages generated by the AI.
Example Prompts
- "Check for new messages from 'Project Team A' and summarize the last 5 entries, ignoring any system join messages."
- "If 'Client-John' sends a file, reply with 'Received, I will review it shortly.'"
- "Scan the current chat window in slow mode and list all participants who have sent messages in the last hour."
Tips & Limitations
- Coordinate Calibration: The script relies on specific screen coordinates (386,68 to 1891,842). If your WeChat window is resized or moved, the 'fast' mode may fail; use the
--slowmode (full-screen capture) as a reliable fallback. - Performance: For high-volume chats, use the default 'fast' mode to reduce overhead. Use
--debugduring setup to ensure the vision model correctly maps your chat bubble colors and positions. - System Messages: The built-in prompt engineering logic is designed to filter out system notifications (e.g., 'User joined the group'), but ensure you test in a quiet environment first.
- Security: Since this interacts with your private messages, ensure your API key remains secure and that you only run the script on trusted devices.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-chenxundaozu-wechat-qwen-reply": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: file-write, file-read, external-api, code-execution