punk2898/wechat-group-stats

WeChat stores your local chat history in an encrypted database file. This Python tool decrypts that database (on macOS only) and produces activity statistics for every member of a group chat, showing how many messages each person has sent in total and over the past one, three, and six months. Members are automatically sorted into activity tiers based on their message frequency: super active, active, occasional, low frequency, dormant, and inactive. A small local web server then displays all of this in a browser-based dashboard with a dark theme. The table is sortable and searchable, and has a refresh button so you can pull fresh data without restarting anything. The analysis can also be exported as a JSON file for use in other scripts or automated tasks. The decryption step relies on a separate open-source project called wechat-decrypt, which extracts the encryption key from WeChat's memory while the app is running. That step requires WeChat to be re-signed with a loose code signature on macOS, which is a one-time setup. After that initial setup, the daily workflow is just running the web server and clicking refresh in the browser. Setup takes several steps: cloning the decryption dependency, compiling a small C program it includes, re-signing the WeChat application, then cloning this project and running the analysis script. The README walks through each command explicitly. The tool is described as intended for personal data only, meaning your own chat history. It works on macOS only because the key extraction step depends on a macOS-specific system interface. If WeChat updates its encryption, the key extraction and re-signing steps need to be repeated. An optional Hermes Agent skill is also included, which can guide a conversational AI agent through the full setup and analysis flow automatically.