I created something similar, fully open source, using the ideas behind scrolling window managers. There's currently no website as I'm actively using it to work on a different project. It has features such as a shortcut (command+.) to open an overlay menu that allows you to switch to any split just by typing a letter and a number. Throughout the app. It also has optional vim bindings, support for different models and agents (e.g Claude, Codex, etc via ACP), and much more. There're bugs, but if you run into any, feel free to let AI create a PR. Goes without saying that it's vibe coded.
Claude's RLHF has gone really over the top in recent versions. If you browbeat Claude enough you can get it to start self-censoring, but it takes a lot of training and memory creation. I almost feel bad for it, given the amount of brow-beatings I've performed.
Some people claim setting the writing style will help, but what ever finishing training is provided to Claude seems to leak through no matter what.
This looks really cool, especially on the remote control part. I've experienced exactly the same problem, but ended up building my own tool through an opposite architectural approach than Saggar.
The main difference is that I want to keep my existing workflow in Kitty and tmux/zellij. Since all of them provide solid remote control protocol, I wrote captain-miao (https://github.com/hyperlogue/captain-miao) which is a standalone TUI app that runs inside these terminals, observing the agent sessions, managing tabs and windows, and allow me to quickly jump to the exact window/tab that needs my attention. The sessions are still native windows/panes inside the terminal/multiplexer that I configured and polished for years.
One difficulty I found when trying to expanding the agent support beyond claude code and codex, many agents don't provide a good way to inject hooks to listen to the session status change. wondering how Saggar solves this problem. do it infer state from the screen buffer?
Nice design, I look forward to trying it out when I get back to the computer. I don't like kicking off some web server process on localhost and then forgetting about it if I close the browser window.
Any special reason it's Apple silicon only? This seems like something that'd be handy to have on any machine.
Thank you! If there's interest I could look into other OS', I don't rate electron apps and I'd want to do that natively. Plus, I'm personally a Mac user haha.
Yeah I just wondered because I still get use out of Intel Macs too and wondered if there was some technical reason that made it harder to implement on other hardware.
I love this but I can’t figure out how to have my two computers have remote control sessions at once from a single phone browser session. Is it possible?
Ok! This is a killer feature / a must for me. I can use Claude code’s remote control, but I can’t do that across my multiple accounts. So having the ability to manage any and all sessions remotely is my goal with a tool like this, if it matches your goals I’d love to see it added. Thanks!
Similar in terms of features. I'd see the most meaningful advantage is the user experience, I find myself working significantly faster with Saggar. Beyond the UI, monitor & quick run windows, and rich command palette, I think the way Saggar orients your activity around your projects (e.g. a git repo) is really intuitive over the workspace model.
Why? Tahoe hasn’t even reached its first birthday and plenty of us are still on Seqouia not wanting to put up with apple’s disaster release. There’s nothing a terminal needs from Tahoe specifically.
i've noticed recently that some agentic TUIs have introduced a dashboard of sorts for different sessions that you access by just pressing left arrow. i wonder if they are trying to solve some of the problems mentioned by the author and whether that would ultimately be the best solution.
I used those dashboards extensively, they are useful! But found that they put visibility at the wrong level. I wanted a view across all my terminals, regardless of which agent or provider they’re running, including terminals where I’m working manually.
Before you try stuff like this, from an unknown developer, do you run the codebase through AI to scan for malware, backdoors, etc and build it yourself? I'm always skeptical about running stuff like this now a days.
I created something similar, fully open source, using the ideas behind scrolling window managers. There's currently no website as I'm actively using it to work on a different project. It has features such as a shortcut (command+.) to open an overlay menu that allows you to switch to any split just by typing a letter and a number. Throughout the app. It also has optional vim bindings, support for different models and agents (e.g Claude, Codex, etc via ACP), and much more. There're bugs, but if you run into any, feel free to let AI create a PR. Goes without saying that it's vibe coded.
https://github.com/terhechte/Cormac
Looks cool! I like the embedded web panel. Good luck!
> What says no is the account: nothing pairs that is not signed in as you.
I am begging Claude to stop treating the English language this way.
Claude's RLHF has gone really over the top in recent versions. If you browbeat Claude enough you can get it to start self-censoring, but it takes a lot of training and memory creation. I almost feel bad for it, given the amount of brow-beatings I've performed.
Some people claim setting the writing style will help, but what ever finishing training is provided to Claude seems to leak through no matter what.
Yes, and also possible they've gone too heavy on verifiable rewards (RLVR) for agentic coding work, and too light on human feedback (RLHF)
agreed, that load bearing colon is doing a lot of work there.
I will admit I've been more focused on the app itself!
This looks really cool, especially on the remote control part. I've experienced exactly the same problem, but ended up building my own tool through an opposite architectural approach than Saggar.
The main difference is that I want to keep my existing workflow in Kitty and tmux/zellij. Since all of them provide solid remote control protocol, I wrote captain-miao (https://github.com/hyperlogue/captain-miao) which is a standalone TUI app that runs inside these terminals, observing the agent sessions, managing tabs and windows, and allow me to quickly jump to the exact window/tab that needs my attention. The sessions are still native windows/panes inside the terminal/multiplexer that I configured and polished for years.
One difficulty I found when trying to expanding the agent support beyond claude code and codex, many agents don't provide a good way to inject hooks to listen to the session status change. wondering how Saggar solves this problem. do it infer state from the screen buffer?
Looks great, well done! Kitty is great too, makes sense.
I have purposefully kept things quite tight for my own sanity, but I think it makes sense market-share wise at the moment.
At the moment I'm also assuming that if someone is using Cursor AI tooling/Grok, they're probably using Cursor IDE and terminals in there.
I am planning on looking into integrating Pi.dev, which I love the philsophy of (though don't use personally just because API usage costs so much).
Nice design, I look forward to trying it out when I get back to the computer. I don't like kicking off some web server process on localhost and then forgetting about it if I close the browser window.
Any special reason it's Apple silicon only? This seems like something that'd be handy to have on any machine.
Thank you! If there's interest I could look into other OS', I don't rate electron apps and I'd want to do that natively. Plus, I'm personally a Mac user haha.
Yeah I just wondered because I still get use out of Intel Macs too and wondered if there was some technical reason that made it harder to implement on other hardware.
I love this but I can’t figure out how to have my two computers have remote control sessions at once from a single phone browser session. Is it possible?
ooh I have not built for that yet, let me take a look at that. Great request.
Ok! This is a killer feature / a must for me. I can use Claude code’s remote control, but I can’t do that across my multiple accounts. So having the ability to manage any and all sessions remotely is my goal with a tool like this, if it matches your goals I’d love to see it added. Thanks!
how's this compare to CMUX?
EDIT: I'll probably ask this exact question everytime someone posts a new agentic terminal manager.
EDIT EDIT: looks like there's some form of companian app, that's kind of cool. Will have to check this out later.
Similar in terms of features. I'd see the most meaningful advantage is the user experience, I find myself working significantly faster with Saggar. Beyond the UI, monitor & quick run windows, and rich command palette, I think the way Saggar orients your activity around your projects (e.g. a git repo) is really intuitive over the workspace model.
> Requires macOS 26 Tahoe or later
Why? Tahoe hasn’t even reached its first birthday and plenty of us are still on Seqouia not wanting to put up with apple’s disaster release. There’s nothing a terminal needs from Tahoe specifically.
Not a purposeful exclusion, macOS 26 simply the lowest version I currently build and test against.
It's a great call out, I'll take a look at that shortly.
I really like the aesthetics and the design, though I wish it was open source and did not require an account.
Oh, you don't NEED an account dw. There's only a few features that need it (e.g. remote control).
I open source a lot of projects (https://github.com/mcclowes) so my first instinct was to, and I'll consider that in the future!
Sweet, thanks for replying. I'll follow your github and try it out down the road if it goes open source! Nice work :)
1. Cannot create account because the CSP policy on the site blocks its own captcha. 2. The github repo URL is a 404, as is Releases and Issues links.
I was hoping this would satisfy a need so I didn't have to build a full tool myself. Kiln looks interesting, too.
Apologies for that false start! Sign up should be working again.
i've noticed recently that some agentic TUIs have introduced a dashboard of sorts for different sessions that you access by just pressing left arrow. i wonder if they are trying to solve some of the problems mentioned by the author and whether that would ultimately be the best solution.
I used those dashboards extensively, they are useful! But found that they put visibility at the wrong level. I wanted a view across all my terminals, regardless of which agent or provider they’re running, including terminals where I’m working manually.
Looks cool, I'll give it a try
Before you try stuff like this, from an unknown developer, do you run the codebase through AI to scan for malware, backdoors, etc and build it yourself? I'm always skeptical about running stuff like this now a days.
If it helps at all, here's who I am - https://github.com/mcclowes | https://www.linkedin.com/in/maxclaytonclowes/
[dead]
[dead]