GitHub · Media and content
jev-chat/jev-chat-windows
The read-the-screen pattern, in a package a non-developer can install.
What it does
Sits beside a chat window on Windows, reads the conversation through local OCR, and proposes three replies for the user to pick from, so the model never writes into the conversation by itself.
The repository's own description: “JevChat-Windows：聊天窗口旁挂的回复辅助。窗口截图 + 本地离线 OCR 读对方消息 → Jev 判断意图 → 3 条候选一键填入，发送永远手动”
How it works
Screenshots are read on the machine, the text goes to Jev with a bounded question about intent and tone, and the answers are rendered as three candidates the person can edit or ignore.
The repository, by the numbers
Read from the GitHub API on 2026-09-24. Stars and the last push move daily — quote them with the date, the way we do.
What we checked
- the README (in Chinese) and the related Android repository it names, read 2026-09-25
- the release link for the downloadable build
- the licence file
What we did not check. Whether the OCR and the model hold up in a real conversation. We did not install the application, and the repository publishes no accuracy figures for either half.