pgszz/wechat-auto-reply-ai

★ 12PythonAudience · developerComplexity · 3/5Setup · hard

Mindmap

mindmap
  root((wechat-auto-reply-ai))
    What it does
      Auto-reply to WeChat
      Keyword rules
      AI fallback replies
    Controls
      Group chat filters
      Cooldown limits
      Mention-only mode
    Web Dashboard
      Start and stop
      Live logs
      Error view
    Setup
      Python 3.13
      Specific WeChat version
      wxauto library

mindmap root((wechat-auto-reply-ai)) What it does Auto-reply to WeChat Keyword rules AI fallback replies Controls Group chat filters Cooldown limits Mention-only mode Web Dashboard Start and stop Live logs Error view Setup Python 3.13 Specific WeChat version wxauto library

Click or tap to explore — scroll the page freely

Things people build with this

USE CASE 1

Auto-reply to WeChat messages from specific contacts using keyword-triggered preset responses.

USE CASE 2

Connect any OpenAI-compatible AI API to generate context-aware replies in WeChat group chats.

USE CASE 3

Set cooldown limits and mention-only rules so a WeChat bot replies selectively without looking suspicious.

Tech stack

Pythonwxauto

Getting it running

Difficulty · hard Time to first run · 1h+

Requires Python 3.13, a specific older WeChat for Windows version, the wxauto library, and the WeChat window must remain visible on screen at all times.

No license information was mentioned in the explanation.

In plain English

This is a Python tool that runs on Windows and automates replies to incoming WeChat messages on a personal WeChat account. WeChat is a widely used messaging app in China. The tool works by keeping the WeChat desktop client open and watching for new messages from contacts or group chats that you configure it to monitor. When a message arrives, the tool checks it against a list of keyword rules you define. If the message contains a matching keyword, it sends back the preset reply for that rule. If no keyword matches, it can fall back to an AI service: you supply an API address, a key, and a model name from any service that uses the same format as OpenAI's API, and the tool calls that service to generate a reply based on the conversation context, including recent message history and the current time. For group chats, there are extra controls. You can tell the tool to only reply when someone mentions you by name, restrict replies to messages from specific senders, or set the AI to judge for itself whether a reply is appropriate rather than responding to every message. A cooldown setting lets you cap how often the tool replies to any single contact or group, which helps avoid looking like a bot or triggering account restrictions. The tool includes a small built-in web interface accessible on your local machine, where you can start or stop the auto-reply function, view live logs of what messages were received and what replies were sent, and check for errors. Setup requires Python 3.13, a specific older version of WeChat for Windows, and the wxauto Python library. The WeChat window must stay visible on screen at all times, minimizing it breaks the automation. The README notes that automating a personal WeChat account carries some account suspension risk and recommends testing with a secondary account.

Copy-paste prompts

Prompt 1

Set up wechat-auto-reply-ai with DeepSeek as the AI provider by filling in the API address, key, and model name in the config, then test it by sending a message from a secondary WeChat account.

Prompt 2

Add a new keyword rule to wechat-auto-reply-ai that replies 'I am away until Monday' to any message containing the word 'urgent'.

Prompt 3

Configure wechat-auto-reply-ai to only reply in a specific group chat when someone @mentions my username, and set a 5-minute cooldown between replies.

Prompt 4

Use the wechat-auto-reply-ai local web dashboard to view live logs and confirm that AI-generated replies are using the correct conversation history.

Open on GitHub → Explain another repo

← pgszz on gitmyhub — every repo by this author, as a profile.

Verify against the repo before relying on details.