Honestly right now I do a lot of development from my phone - I don't have the time to pull out my laptop very much. Right now I use hermes agent and just chat with it on Slack. But sometimes I wish I had easier access to more advanced features, and Slack isn't the best UI for this stuff. So I want to use something like this but am having analysis paralysis.
If you’re happy with OpenCode as the harness, OpenChamber is great. But if you prefer using different harnesses under the hood, Paseo is a better fit.
I installed it on my homelab and can access the same sessions from my MacBook or iPhone at any time. Been really nice to use so far.
are people using these things but isolating agents to some sort of containerization individually or do you have like one big dev env ? probably time i get around to sorting a better dev container strategy just to isolate agents here but trying to think of the best way to set this up locally.
I really think that using something like Paseo (I use Ouijit[1] because I think it's actually a step ahead of where most of these multiplexers are) and nono[2] is basically the best coverage you can have, and really lets you go crazy with auto mode or permission skipping withiout having to worry about any security concerns.
A lot of people lean into the VMs but realistically it's not actually doing a lot of the security work that you need and leaves you wide open to prompt injection/exfiltration attaacks. The flexibility of a good nono setup (which is easy because they ship a nono skill that claude can consume) actually locks things down significantly better
This actual functionality should be more explicitly stated, and should be described much sooner IMO.
[1] https://air.dev/
I have a number of these pieces in a binary I drop first thing on every agent computer: https://github.com/housecat-inc/scratch
Chat to code with an inspector to point at specific DOM sections, review diffs on the box, hand off to mobile, etc.
Makes me wonder where all this will end up. Or DIY win with hyper-personalized tools? Will one vendor become the center of gravity similar to VSCode in previous generation? Will there be a cottage industry of more opinionated tools?
What do you when done with a work tree? Just delete? I wish there was a middle ground of hiding it from the main panel, but keeping it for a week or so just in case I need to refer to something in the transcript...
You can "sleep" the worktree so that it shuts down the terminals / agents to not consume RAM. Then in the worktree card filter, you can select "hide sleep"
Later you can "unhide sleep" and click into the worktree to wake it up if you need it. It will restore all the agents
Any suggestions for terminal UI versions of something like this?
First screenshot is of a phone.
Yeah I’m never installing a coding agent on my phone.
Both codex and Claude have remote from phone. I’m personally using cli + terminal from phone.
I’m sure I’m not the only one.
so far it's been working pretty well for me in most cases, but not always. having a few parallel efforts in motion means that even if one goes off the rails, i can pause it and still make progress on the others, then fix it up when i get home.
It’s not great but it’s better than anything else I’ve tried.
They fumbled the initial rollout with the way they gate-kept the agents behind their own subscription. I used Claude through its API pricing for a while, and the UI experience with the sidebar chat was so much better than the Claude CLI.
I could see the thought stream, click on artifacts, view diffs, etc. But it burned through tokens way too quickly, so I downgraded to the CLI agent.
At least I can use IDEA's inspections through the MCP server. It enables the agent to use the IDEA's index for navigation (finding usages, locating symbols, etc.)
I've said it before and I'll keep on saying it: the churn in new dev interfaces it's super exciting. Can't wait to see what next week brings.
As a simpler desktop-only multi-platform option, I can highly recommend Termic (https://github.com/simion/termic). It just runs existing TUI coding agents (so Codex, Claude Code, OpenCode, etc) and manages them in sandboxed git worktrees. You get a real PTY to manage agent in, lets you see when one is working or has paused, plus a terminal, "changed git files", even in-line-code comments you can feed back to the agent. It's a good design, works well, the main dev is working very quickly on improvements, and they're responsive to issues and PRs.
OpenChamber, OpenAI, Open WebUI, OpenDoor, Open weights, OpenTable, OpenClaw, Open This, Open That. We're increasingly diluting the meaning of the term, and "open source" more broadly, that we're at the point of OpenFatigue being a real concern here.
OpenConcern.
But at least it's better than Concern.net or iConcern, or even eConcern.
Let's not forget Digital Concern, ConcernCloud. And my least favourite, Concern On BlockChain.
MIT license.