hob - The professional workspace for your whole agent stack
by•
Run complex agent work while staying in control. hob brings multiple models and independent workflows into one professional workspace. Agents can shape the workspace around each task, coordinate through it, and guide you inside it. Run, review, automate, and recover agent work in the same system, so you can direct more parallel work without piecing together the infrastructure.
Replies
hob
Hey Product Hunt 👋
I’m Andrew, one of the people building hob.
We built hob because coding agents changed the job - but the tools surrounding them are still extremely fragmented.
As you use more agents, your stack keeps growing: separate terminals, worktrees, reviews, automations, remote environments, and session histories. You become the connector between all of them - moving context around, tracking what each agent produced, and reconstructing why decisions were made.
hob turns that growing stack into one professional workspace.
But hob isn’t simply a place where you can see multiple agents. The agents themselves are users of the workspace - so you and your agent team are always on the same page.
They can shape it around the task, open the context they need, coordinate work through it, recover previous sessions, and guide you inside the app. The workspace becomes part of what your agents can do, so you don’t have to manually connect every step of their work.
Claude Code, Codex, and OpenCode can work alongside terminals, isolated worktrees, pull requests, automations, artifacts, remote access, and headless mode. Commits, issues, and plans remain linked to the exact conversation and turn that produced them.
We also built hob to protect your interests. Your provider subscriptions run directly inside hob, and we don’t sell inference or resell tokens. Your workspace runs on your machine, with end-to-end encryption when you access it remotely.
Download hob and try it free for 1 month. There's a button in the app to start a trial with no credit care required!
https://hob.dev/download
If it works for you, the first 1,000 personal subscribers can lock in 40% off for the life of their continuous subscription.
We’d genuinely love to hear what your current agent stack looks like, and what you’d want your agents to handle themselves inside hob.
Openfort
Awesome video btw! got my hob :)
hob
@joalavedra Welcome to the hob community! Let us know how your experience goes!
Subanana
Woah! The video was amazing. Also, love the idea of workspace for the agents. Congrats on the launch @callmegwei @chetrit_ @josepchetrit
hob
@lakshya_singh The video tries to capture the feeling of using hob... you're not tethered to your desk, you're not switching apps all day, work becomes more productive but more mellow, controlled, and organized all at once. Hope that landed!
Take hob for a spin and let us know if its an accurate portrayal! 🏖️
Does this work with all agents and LLMs? How many workspaces can you have at one time? I tend to have multiple projects on the go at once so being able to have one unified place where this can be viewed, managed, and coordinated would be valuable. Loom forward to giving it a try.
the constraint that bites first is not infrastructure, it is review budget. running ten agents in parallel is easy now, and reading ten outputs carefully is the same afternoon it always was. so the part i would want hob to be good at is review. running more of them is already solved by anyone with a terminal.
hob
@rabnoor_s Right, its not just running them - its all the stuff that happens post run. What agent did the work (commits linked back to conversations), what context is going to help with the new issues the agents spawned along the way (issues linked to created conversation blocks) - and how you review that work afterwards (no leaving hob to review and make a few manual tweaks to the code changes, the images generations, the markdown docs, etc). Observability and organization are critical - not just while your ten agents are running, but before and after.
oh nice timing. i've got a few claude code agents going in parallel worktrees and after a day i genuinely lose track of which one did what. commits linked back to their session, um that'd help a lot.
quick one, does each session get its own worktree or do tey all share the same tree?
Congrats on the launch btw 👍
hob
@seenew You can create each workspace either as a new worktree or just as a grouping of panes that exist inside the project and share a branch. So you get the best of both worlds. Sometimes worktrees make sense, sometimes they are overkill - so we accommodate both.