OpenChamber: An Agentic Development Environment
openchamber.dev
openchamber.dev
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.
Worktrees are cheap because they share the object store, so N agents each get their own working dir + branch without N full clones, and teardown is just `git worktree remove`. That solves the common case — two agents stomping each other's uncommitted files on the same repo. What it does NOT give you is isolation of the filesystem outside the repo, the network, or your secrets: a worktree agent can still read ~/.aws or curl anywhere.
So I only reach for a container/VM the moment an agent runs untrusted code or I care about blast radius — npm install on a sketchy dep, executing model-written shell, anything that could exfiltrate env vars. Rule of thumb: worktree for "don't let agents collide", container for "don't let an agent hurt the host". Most local setups only need the first, and people tend to reach for heavyweight sandboxing before they need it, then abandon it because it's painful.
A cheap middle ground that works locally: worktrees + a per-agent env file and a restricted PATH, so at least secrets aren't ambient without paying full container cost.
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
> ACP catalog, any agent speaking the Agent Client Protocol is supported through a generic adapter. Paseo ships a curated catalog of one-click installs (Cursor, Gemini, GitHub Copilot, Hermes, Kimi, Qwen Code, and 25+ more), and you can add any other ACP agent yourself.
> Either way, you install the underlying CLI. Paseo runs it.
That is the value. You can run them remotely from your phone and set them up on a headless machine too. You’re not bound to your terminal.
My usual workflow is to still use the terminal for heavy work. But sometimes I just want to prompt from my phone and run a few tasks or do some planning. Then I can simply resume the session in the terminal later. It's the flexibility.
This actual functionality should be more explicitly stated, and should be described much sooner IMO.
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
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.)
Now, you can use a standard Codex (and maybe Claude) subscription inside of a sidebar, and get a lot of the features you mentioned. It's nice, if you're trying to be more involved with the code, but not as nice for multi-agent stuff.
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.
Features we are proud of are one click onboarding of new VPSs into the agent fleet. Agent orchestration across different machines and harnesses. As an example, a Codex coordinator on one machine can run a Grok implementer subagent on another machine and they can talk to each other.
It's not really ready for public consumption yet, but it's been our exclusive daily driver for the past month. We'd be happy for people to test drive it and provide feeback regardless: https://github.com/madeinorbit/podium
Any suggestions for terminal UI versions of something like this?
Thankfully most of them can be proper Web with my already installed browser, or IDE based.
VSCode plugins get a discount, because I cannot avoid it in some cases.
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?
[1] https://air.dev/
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.
It’s not great but it’s better than anything else I’ve tried.
Tailscale integration works well too. The mobile notification setup was a little clunky and the instructions were confusing, until I found the directory with the shell scripts. It should probably expose a herdr action to generate the keys and include that in the documentation.
One thing though: the subagent lines should be clickable like they are in Termux but your actions box absorbs the click and focuses the GUI input line.
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.
Bring tethered to work 24/7 isn’t freedom from work, it’s enslavement to it.
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.
Makes it feel like actual effort went into this. :)
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.
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.