Connecting the machines

ahknight

Why would I want a service for this when I could easily run such a discovery service myself? Why not just release that tool to users? SyncThing offers its discovery server for self-hosting as an example. Oh, right, this is how we monetize in 2026. Nevermind.

ActionHank

I am not sure I fully understand the need to farm out sufficient work that I need a network of machines to do it. Unless you're just slop farming.

hext

This is fantastic to see, I've been daily driving herdr at work and at home for a few months now. It's great. My biggest complaint has been exactly this, I want all of my ssh sessions available in one spot. Excited to try it out.

CharlesW

I'm hoping someone can explain why this is interesting. I get that Herdr wants investors to give them money to own developers' coding sessions, but the value proposition ("Walk away and they keep working. Come back from any machine and they're where you left them.") seems like a miss since it's so easy to do that without them. I'm surely missing something.

evil-olive

> But until now, each machine needed a separate Herdr client, usually in a different terminal tab. You had to remember where each agent was running and switch between those tabs to check on them. That was the bottleneck. is it, though? I can imagine having a local Herdr instance and one running on a remote box somewhere, but I can't think of a use case where I'm trying to juggle N different Herdr instances and that overhead becomes a meaningful bottleneck. > Getting to 1.0 > After this update, what I want is to make connecting those machines more convenient. You shouldn’t need to think about SSH or deal with complicated network setup. It should be easy to connect any machine, anywhere, through one Herdr account. (emphasis added) sigh. I'm a daily user of Herdr. I don't have a Herdr account today. I don't want a Herdr account. I don't need a Herdr account. the product is not even to 1.0 yet and already the writing is on the wall for the path towards enshittification & acquisition by $bigco. of course, I'm sure there will be some "usage without an account will still work" platitudes, but the trend of development is clearly going to be in the direction of features that "integrate" with the "developers using Herdr pay us money and/or we collect their data" business model.

evanpurkhiser

I’ve been running codex’s remote app server as a systemd service for the last several months, and before that opencode remote. It’s very easy to switch between my phone and desktop app. I haven’t tried herdr yet, but maybe it’s just not targeted at me? The real problem I have with the remote codex setup is when it needs to authenticate as me. I sort of solved this for sudo and ssh with this [1], but I think something like this could be integrated deeper into the harness [1]: https://github.com/evanpurkhiser/agent-witness

bibstha

Worth giving onorca.dev a try. That already supports this with nice UI. Also atcyrus.com is interesting to delegate tasks too.

rvz

Again it bears repeating [0] that if you don't pay for your tools and expect open source maintainers to do it for free, then they will most certainly take funding from VCs and angels instead. Now they have raised $6M in funding. [1] Do not be surprised to realize that open source does not pay their bills. [0] https://news.ycombinator.com/item?id=49201743 [1] https://herdr.dev/blog/herdr-raised-a-seed/

qingshan3

I have been running herdr since yesterday, and have been using this feature to connect multiple machines. Now I want to delegate most of tmux keys to herdr for consistent experience.

bigcat12345678

Makes sense. I had always been curious on why Tailscale hasn't creating agent-oriented features. It's obvious to me that Tailscale is at the best position (AFAIK) to capitalize on agents demand. Their connectivity product is unique. And it's a natural fit to connecting long-running agents instance across the Internet.

cobbzilla

Sorry but whenever I hear the name of this tech I visualize the stereotypical “dumb person” saying “hurr, durr” nonsense words

kelvinjps10

It would be good if they released a good mobile app I'm have been using termux and moshi and claude code. Termux and moshi provide the full functionality of the terminal but it gets clunky sometimes and with claude code you don't get the full functionality and a lot of stuff are missing from the tui

theptip

It’s a good feature. However I am not convinced that the terminal/ssh abstraction will remain optimal when you orchestrate a swarm of workers across multiple machines. Running via ssh is great when your machines are pets that have local state. I don’t find it to be that useful when I have more than two machines that I’m load balancing agent sessions across. Just for example, a nice feature from the Codex app; if you edit a markdown file on your remote, and click on the link to the edited .md file, it just opens the file in the local app, and you can view the rendered form too. If I’m working on my local machine this isn’t a win vs. opening the file in Obsidian / IDE. But it’s way easier than manually viewing the file over SSH (unless you just want to look at the raw text version). Similar applies to html files. I think it’s more likely the Fly Sprites or Claude/Codex sandbox model will win as we move to swarms. If I’m not watching the actual harness output, there’s not a great reason to attach a pty.