Herdr: Agent multiplexer that lives in your terminal (github.com)

Just use tmux no?

does it support a setup where each agent can be in a different SSH session? or must they all run in the same place.

it seems to support running a remote herdr over SSH but unclear if it can add remote agents (each agent has its own sandbox where its installed and you first SSH into it and then start the agent there)

I need a shepherd for my terminals.

Tried this for a couple days, but conductor.build is way better IME. Running _in_ the terminal is a flex, but doesn't actually bring any advantage.

nice, not dissimilar to what i built for https://github.com/storozhenko98/beehive

How do people use terminal multiplexers together with vim?

Ctrl+B is so hardwired in my fingers for scrolling back one screen that there's no way I'm remapping that one in vim itself. So then you have to remap that in your terminal multiplexer, while at the same time there's a bunch of people saying never change the leader key...

Curious what vim users especially do about this?

This is way to complex... Why don't just use some harness which manages all that and give u a good UI?

At least one of us needs to up our dosage.

The utility is not obvious, but I've been using herdr for a few weeks after a friend recommended it, and it's been a great tool.

I had about a dozen different terminal windows, each one with multiple tabs, which was becoming a mess to manage; multiple agents and harnesses running that I could only inspect by remote access/VNC away from home.

With this, I can keep things more organized (project workspaces, tabs), because it's backed by a persistent process, I can ssh into the machine, with tailscale, run herdr and see all active sessions, do some debugging and one-off prompts.

Even better, I can do that from my phone and iPad using Prompt/Termius. It gives me the best part of 'xxclaw' harnesses with none of the complexity.

Will people really need a bunch of different agents? Even using two codex session for different project make me feel a bit overwhelmed, or may be i just old

How does it compare with https://github.com/agent-of-empires/agent-of-empires

I switched from tmux to Zellij a while back, and lately added a stop hook that sends a terminal ping to the correct tab when the agent is done (and I'm not looking at the tab). It has been pretty convenient so far.

Added an example here:

https://johnny.chadda.se/zellij-stop-hook/

[flagged]

How do you use multiple agents in one project? Isn't there race conditions? or is this for 2 separate projects at once?

Once you get it set up, Emacs is a pretty damn good "agent" multiplexer as well. I use agent-shell with Projectile on Doom Emacs as my main workflow these days, and it works very well even if I have 6 projects open or whatever.

Claude and Codex are also both quite good at working with Emacs as well. Depending on your isolation/sandboxing strategy, they can either run commands against your session via emacsclient (a bit scary) or dump elisp in the REPL for you to evaluate. Both are really efficient in terms of giving you a fast feedback loop.

Getting a lot of flack in all these comments, I really like this tool. Have been super easy to use and scale to multiple agents. I've had a ton of issues with tmux and copy and paste, this just works. I was using warp terminal, and even working on my own fork of it with it's recent open source status, but this has won me over.

I started building my own version of this before I discovered herdr, and while its not quite the same it's improved my development workflow by an order of magnitude. https://jmux.build

I am going to give them a try side by side!

Multiplexing the agents is clearly the first obvious pain point, but the other one I keep encountering after this is visibility: with multiple agents running, it becomes difficult to see what each of them is doing, what program did they call, and where they are getting stuck until they complete or fail.

Is there any information from Herdr about what each agent is up to beyond the output? Or does it just concentrate on orchestration for the time being?

Herdr is great, I've been using it for over a month now. Their API is awesome too - you can build tools around it.

[flagged]

if you prefer to self host but with a gui there is https://github.com/kdlbs/kandev

Unlike Cmux, Herder will not call out to multiple web services upon start-up.

has anybody figured out how to get this to open a fixed layout? I'm using tmuxp to do this easily with tmux but would switch to this pretty quickly if it was easy to get my layout over. It's really nice. It's like zellij without the bloat (but with different ai bloat that doesn't seem to drag down performance at all and isn't in your face unless you want it in your face)

this works so seamlessesly inside emacs out the box without a dedicated app

https://github.com/elij/macher-agent

I've been using this and it's been really nice for a specific use case: running agents locally alongside agents in a remote sandbox environment. Most apps managing this split seem to want to manage your sandbox for you, but we've already built out some nice tooling and infrastructure for that that meets our needs. I don't want to add a new vendor in the mix for our infra that may not exist in a year and that will be harder to integrate with. Combined with some additional tooling I made, herdr lets me easily connect to remote agent sessions, see their status in the sidebar and fluidly switch back to work on my local machine from one interface.