Claude Code Desktop App Redesigned - Run parallel coding agents from one desktop workspace

the summary view mode is the most underrated part of this.  been running claude code for months — verbose output is the constant friction. you end up reading claude's narration instead of the diff. having that as a first-class toggle at the session level, not a system prompt hack, changes the actual workflow.

Using this for MentionFox and FoxAPIs right now and I can say it is a bit unstable, but usable. I have Claude managing it for Claude Code automation. I can say that you have to watch Claude code when automated. It will tell you things are done, but it didn't commit. Buttons that do nothing, pages not JSON'd together...

Will there be a Linux version? 🤞

awesome!!

useful stuff ! i like it , what programing language is it native in?

I've been using primarily Claude Code CLI so this will be very interesting to try out later this week. Having everything visible in one window sounds like a treat. I might find I needed this more than I could have anticipated, as someone that's used to the CLI and Linux Terminal I look forward to exploring this. Thank you

Just upgraded @Claude Code Desktop App Redesigned v1.3036.0. If you set any

@Claude by Anthropic env vars for terminal use (API keys, model overrides, custom base URLs), the Desktop app will pick them up and break with cryptic 400 errors.

Fix: move terminal-only vars to .zshrc (interactive shells only) and clear the macOS launchd cache:

launchctl setenv YOUR_VAR_NAME ""  

Then Cmd+Q and relaunch.

Fixed the env bug in 15 minutes. Upgraded to Opus 4.7. Back to building.

This is interesting, can I use it on the MAC? Is there a display option for UI config instead of CLI?

Is this a blog post title or a product launch? who cares?