WeChat

WeChat MCP Server implements the MCP protocol to automate sending WeChat messages from an AI assistant. It exposes tools that let you send text messages to a contact or group and schedule messages for later delivery. After starting the server, your MCP client (AI assistant) can discover the available tools via tools/list and invoke them using tools/call. A typical use case is instructing your AI to send a message to a specific contact or group, or to schedule a message to be sent after a delay. The server handles JSON-RPC 2.0 requests and returns results indicating success or any errors encountered during message delivery.

To use from your AI assistant, register the WeChat MCP server configuration in your assistant’s MCP settings. Then call the available tools by name with appropriate arguments, for example send_wechat_message or schedule_wechat_message, supplying the contact_name and message (and delay for scheduling).

Prerequisites:

• Python 3.8 or newer
• A running WeChat client on Windows (WeChat NT framework support as described in the docs)
• Internet access to install dependencies

Install steps:

1. Clone the repository git clone https://github.com/1052666/WeChat-MCP-Server.git cd WeChat-MCP-Server

2. Install dependencies pip install -r requirements.txt

3. Ensure WeChat is running and you are logged in

   • Open WeChat client and verify you can send messages manually

4. Configure the MCP client within your AI assistant

   • See the example in the README under "配置AI助手" and adapt path if needed

5. Run the MCP server

   • From project root, run: python src/mcp_server.py

6. (Optional) Verify the server is reachable

   • Use your MCP client to call tools/list to confirm available tools are exposed

Tips and notes:

• The server currently supports two tools: send_wechat_message and schedule_wechat_message. Ensure the contact_name exactly matches a chat in WeChat (e.g., a contact or group name).
• The project uses pyautogui and win32gui for UI automation; ensure your Windows environment permits screen control permissions.
• The MCP configuration example uses a placeholder path; replace it with your actual repository path when embedding in your AI assistant config.
• The README notes that Mandarin Chinese messages may cause encoding issues; the current capabilities emphasize English, numbers, and emojis. Test carefully with your target content.
• Do not manually interact with the WeChat window while the MCP server is controlling it for reliable operation.