From early on, I favor AI CLI using terminal over AI in IDE. A session starts blank, and grows as needed. However, even though I like it very much, terminal can be limiting, and agents can do more for us than standard terminal allows. Can it be better? In my spare time, I have been experimenting and knocking off pain points one by one. Now, I have a terminal but not just a terminal, which I use daily for all my projects. I am sharing it open and hope you find it useful. Happy to answer anything.
Show HN: The terminal I live in all day: comment on anything coding agents print
github.com