Somehow reminds me of JetBrain's new Air [1] platform (still in beta I think) - make multiple agents working, monitoring, basic code and diff viewing.
[1] https://air.dev/
Looks nice.
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?
is agent orchestration somewhere in your backlog?
I really enjoy the sidebar style for organising work, I’m not sure why the OpenCode desktop app moved to browser-like tabs (horizontal ones, though tbh my browsers now have vertical ones ehh).
This appears to be somewhat like Paseo, specialised for OpenCode under the hood?
Try codecast for a better version of this that works across all agent harnesses in one unified way.
It's nice but I get such memory leaks that I have to reboot my bacbook after a few hours of using it. Sticking to opencode gui for now.
It's funny / ironic how the whole agentic development boom has taken off in terminal environments, but now the orchestration / IDE layer seems to be all Electron / GUI apps.
Any suggestions for terminal UI versions of something like this?
It feels a bit like herdr which is like tmux meets agent management
Had to scroll all the way to the bottom to see that its a wrapper for OpenCode.
This actual functionality should be more explicitly stated, and should be described much sooner IMO.
on the scroll note, the top background had a seizure while i was scrolling down
Over 50 npm dependencies (dev and non-dev). Is this really how things are developed in 2026? After all what's happening (cyber security), is any developer out there running this kind of software just like that in their main machines without any concerns?
It is if someone forgot to prompt their harness "don't pull in unnecessary dependencies".
I'm using Orca recently for both coding and as a terminal and like it a lot, seems this is very similar, just tied to a single harness?
Huge fan of Orca. Been using it for a month. Plan to setup a proper remote env this week.
Orca is my main ADE too, but I haven't found any other users!
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...
Even though all these "ADE"s are mostly the exact same (Superset, T3, Conductor, Paseo,random Reddit-dude,etc.). I'm liking Orca largely because of the built in browser with ability to select an object(s), write a comment and send it back to the agent. I had built my own solution, but Orca makes it even easier and convenient.
Huge fan of Orca too.
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
Why use this over T3 Code?
OpenEverything
Sticky header with no background seems a bit awkward on mobile. The text overlaps.
From now on I swear if I launch a new product I will call it ClosedXXX
[flagged]
Seems like another clone of Conductor, though this one has a serious amount of feature creep (supporting so many platforms will add a ton of extra complexity), and on OpenCode SDK only.
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.
Is saying "XXX: An Agentic Development Environment" already sufficient to get to the front page of HN? I guess no differentiation is even required in this category, huh?
Looks interesting, might try it out. I have to ask, did you really need to hijack my scroll behaviour?
This looks pretty awesome, like almost exactly what I've been hoping for. Is it possible to run the sessions on a remote server to keep tasks running? I mostly work from my laptop and it's nice to fire off tasks to a remote server (mostly been Cursor Cloud Agents lately) and close my laptop, maybe check in from the app later.
Yes
I have been using Opencode desktop, going to try out these other tools too
Since this page is clearly made with AI, someone deliberately prompted to add scroll hijacking for some reason, despite it being known to be bad UX.
This looks neat. How is this different from Jetbrains Air?
Oh, I love the idea of a changes walkthrough. Maybe I'm old fashioned but I still look through all the code and prose that gets committed. In Orca, it took me about half an hour to even find the diff view (hidden behind a panel, activated on the top right)
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.
[dead]
I wonder what JetBrains is doing. They had such a strong lead, but with the rise of AI, they seem to be heading down the same path as Stack Overflow.
I've been paying JetBrains in a personal capacity for about 15 years. Almost from the first year of my professional career. I didn't renew this year. I feel so sad and disheartened that a company that produced such brilliant tools have completely lost the plot on AI. They had started work on Fleet, a lightweight VSCode like IDE, before LLMs became big, but abandoned the effort without seeing it through. It could've been a great agentic IDE (all I want is whole file+project view for every conversation, instead of just diff so I can have a better perspective of the codebase). Instead they built multiple confusing tools (might be internal org structure and competition leaking out), and took far too long to adopt SOTA CLI coding agents, and kept charging for it as a separate subscription. I wish they do a whole company rethink on agentic development, but it is going to be tough because the existing enterprise licenses will continue to be the main revenue driver for a few more years, until it all dries up all at once.
Yes, I haven't heard as much about them lately as I had in the 2022-23 era.
I use IntelliJ at my current gig—and the CoPilot plugin for AI, because it’s what I have access to—and it has become ClaudeCode-esque of late: sub-agents, skills, tool-calling, hooks…and continues to improve. It’s unfortunate how the AI-in-a-sidebar UI kind of hides how sophisticated it has grown, but it’s a glass of ice water in hell in an environment where you can’t easily procure what you want.
[dead]
They have the Air and it's quite nic for agentic coding.
I let my Pycharm sub lapse since I switched to VSCode. Its lighter and better at some static analysis.
Same thing. I've been using JetBrains IDEs for 25 years, and by now the keyboard shortcuts are hardwired into my fingers.
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.)
same, but I upgraded from the in-IDE AI to claude code :)
I prefer Paseo https://github.com/getpaseo/paseo, mainly because I have specific preferences for different harness + model combinations. For example, I like using ChatGPT models via pi, and GLM via Claude Code.
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.
+1 for paseo. Switched yesterday and it’s a dream how flexible and clean it is.
+1 for Paseo. Switched over to it a few weeks ago from conductor.build then superset.sh. The mobile app plus Hetzner VPS has been really enjoyable in terms of just firing off random tasks and not being stuck at my desk.
As someone who uses both Claude Code and Codex, having a unified interface while still being able to use their respective harnesses is a godsend. I've tried a lot of alternatives, and nothing even remotely comes close to the level of polish this one has. The session sync is buttery smooth as well, it's even better than Claude Code mobile I would say. Thanks so much for the recommendation.
this looks prettttttty slick i might dive in later.
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 had another comment in a thread last week about this: https://news.ycombinator.com/item?id=49111625
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
It’s just a personal preference, to be honest, so I wouldn’t overengineer it yet. In my case, I run it in a VM behind Tailscale. The main reason is that I want it to be able to install or use whatever tools it needs without contaminating my main machine. It can expose anything I need over Tailscale. So for me, it’s basically just a cheap, throwaway machine.
[flagged]
I have a VM in which I run my general compute. In that VM I use incus linux container and run these tools as unauthorized user of the container. That should be reasonable security while having nice control for resources.
This looks very nice and well done. Good job! Okay, I'm going to have a tongue-in-cheek rant now.
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.
Don‘t forget Open Closed.
Specifically about OpenChamber: https://github.com/openchamber/openchamber
MIT license.
OpenSerious about these OpenStuff claiming to be Open but they are openly not. Looking at you OpenAI
Increasingly? It sounds like you've never heard of the Open Group or open architecture. Some of the oldest uses are some of the weakest.
[dead]
>> "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.
Damnit! If I hadn't already blown $170 locking down `openfatigue.ai` today I would snap that one up too!
> development environment
First screenshot is of a phone.
Yeah I’m never installing a coding agent on my phone.
I frequently use openchamber on the go. It's very good at this.
You can also use it as a plain old chat client. I prefer it over OpenWebui because I can actually carry over my OpenCode configuration, which has specialized subagent definitions (for example, the "research" agent fires off a bunch of "web search" agents that use a cheap model to fetch and summarize, and the more capable "research" agent pulls it all together, decides if more search effort is necessary, etc.). So I can VPN into my home network and boom, self-hosted chat interface that's more capable than what you can (easily) do in OpenWebui.
Why not?
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.
Tiny screen that can barely show any information and the awkward miniature keyboard that's slow to use and gives no feedback, it's a pretty awful way to do anything.
lately, i like to try to spend the morning talking through and framing up a few very detailed initiatives for parallel efforts that each spawn and orchestrate their own tasks, then feed them to /goals and go for a nice walk in the park, get some groceries, check in on the orchestrators and spawned tasks along the way, steer them from the phone.
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.
no issues here with a small iPhone and Blink. I can change the text size on the fly, and am very used to typing on the phone keyboard. Also use dictation
Because I’m not putting work in my pocket.
I don't use it for work I use it for my private projects and this allows me to work on them even when I'm commuting which used ti be wasted time.
id rather have it in my pocket and get stuff done while im out doing what I feel like doing instead of sitting in an office all day.
Are there any Android remote terminal apps that don't suck on mobile? Termius is painful.
if you're using herdr, i built a plugin exactly for this reason. its a pwa so requires a proper frontdoor for secured access (tailscale, netbird etc.).
I’m using Termius.
It’s not great but it’s better than anything else I’ve tried.
It doesn't work with tailscale. Won't prompt you to go through the web sign in.