What is wechat-radar?

WeChat Radar is a local-first dashboard for people who are members of many WeChat group chats and find it hard to keep up with what matters. Instead of scrolling through a long message history, the app reads your local WeChat data on your Mac and turns it into a daily briefing: what was discussed today, which links were shared (deduplicated and given readable titles), messages where you were mentioned, and a summary report for each active group. The app runs entirely on your own computer. Chat data is stored in a local SQLite database file under a folder in your home directory, and nothing is sent to a third-party service. The README includes a clear privacy note: using this tool is your responsibility in terms of complying with WeChat's client rules, local laws, and the privacy expectations of the people in your groups. One feature is a topic radar that clusters messages from across all your groups by theme for a given day. This is powered by OpenAI's Codex CLI (a separate tool) if you have it installed, without it the dashboard still works but topic grouping is less sophisticated. The per-group daily report is designed to be copied and pasted into an AI assistant for further processing, such as drafting replies or summarizing discussions. The setup requires macOS with WeChat 4.x already logged in, plus a separate command-line tool called wx-cli that handles reading the local WeChat data. You also need Node.js 20 or newer and the pnpm package manager. After a one-time setup step at /setup in the browser, you select a date range and click rescan to pull in messages. A demo mode is available if you want to try the interface without connecting to real chat data. The project is licensed under MIT and is built with Next.js, ECharts (for charts), and better-sqlite3.