I sometimes use 5 agents at once, but 3 work on separate tasks and the 4th is watching them by tmux and steering, and the fifth is also watching the whole team for testing quality standards. So 2 out of five are just helping me watch the other 3.
But this only works because the 3 worker agents have really separate problems, almost no connections across.
I dont care about codex,cc harness. but I do miss a proper terminal in pi web; I will likely endup forking it... the lamma cpp built-in web app is super nice, but also needs to be hacked to support api providers, and also no terminal :/ .
I think the definitive solution way would be using vscode stripped for web (what gh codespaces use) and add pi to it.
Looking at the screenshots, this looks close to what an agent factory for teams could look.
Agents on multiple servers, people can see each conversation. Test it…
an ide built around agents instead of files is the direction i keep expecting to win. how do you handle state when several agents touch the same repo at once
Worktrees or sandboxes. You can tell Claude to put its stuff in a worktree. I can easily develop any amount of features in the same codebase because they all run in their own worktree
I went from IDE (VS Code) to terminal (wezterm) running half a dozen agents in an app called Herdr. My computer work is mostly in the terminal now, prompting agents. And they have access to my Obsidian vaults. Can read, and add to them. And access all my notes via a local RAG. I neither need projects in Claude, nor IDEs. Just agents accessing my local vaults and code repos.
If you're more of a SWE than a data scientist, check out Mjolnir: https://github.com/BrokkAi/mjolnir/
Built in containerization and EC2 support, allows moving sessions across harnesses/profiles/machines
Goodness - triggers my ADHD. Power to those who find it useful, I would be so lost in the wrong details (the quota screen is pretty cool).
I sometimes use 5 agents at once, but 3 work on separate tasks and the 4th is watching them by tmux and steering, and the fifth is also watching the whole team for testing quality standards. So 2 out of five are just helping me watch the other 3.
But this only works because the 3 worker agents have really separate problems, almost no connections across.
I feel you! And the watch agent sounds like a good approach
It's a 2 day old project. https://github.com/ZhengyiLuo/AgentsDock
That is new repo for open source, we've actually worked on it for a while
Noted. For the rest of the internet, I was wrong. Here is the evidence. https://github.com/ZhengyiLuo/AgentsDock-Releases/releases
That is "from last year" in human time.
Created a new repo to open-source since the old one has too many commits and some uncessary histories haha
I dont care about codex,cc harness. but I do miss a proper terminal in pi web; I will likely endup forking it... the lamma cpp built-in web app is super nice, but also needs to be hacked to support api providers, and also no terminal :/ .
I think the definitive solution way would be using vscode stripped for web (what gh codespaces use) and add pi to it.
IDEs are today's JS frameworks
Isn’t this just paseo.sh with less features and tightly-coupled tailscale?
Looks useful! curious how you would compare it against something like Paseo?
Tried it out, missing a few important features we need : ). Paseo is great though
What features I paseo missing that you have?
Looking at the screenshots, this looks close to what an agent factory for teams could look. Agents on multiple servers, people can see each conversation. Test it…
Let me know how it goes!
an ide built around agents instead of files is the direction i keep expecting to win. how do you handle state when several agents touch the same repo at once
Worktrees or sandboxes. You can tell Claude to put its stuff in a worktree. I can easily develop any amount of features in the same codebase because they all run in their own worktree
Yeah, and we don't limit how you do it — each chat gets its own tmux session, so it is up to you on how to handle it
I went from IDE (VS Code) to terminal (wezterm) running half a dozen agents in an app called Herdr. My computer work is mostly in the terminal now, prompting agents. And they have access to my Obsidian vaults. Can read, and add to them. And access all my notes via a local RAG. I neither need projects in Claude, nor IDEs. Just agents accessing my local vaults and code repos.
Yes, we need agent-oriented IDE not the old VS code style!
I suggest putting the tests in a separate folder ;)
hhh you spotted my messy folders ;)
looks interesting! Trying it out. Thanks
- arent there 10000 projects that do this
- what makes this one special?
Geared toward for AI research so it is has a lot of nice-to-have features.
Trying this now
Thanks!! Let me know if you need any help