Launching today
Baton
Orchestrate your AI coding agents
58 followers
Orchestrate your AI coding agents
58 followers
Baton is a desktop app for developing with AI coding agents. Run multiple agents in parallel, each in their own git-isolated workspace. Works with Claude Code, Codex, OpenCode, and any terminal-based agent. Smart notification badges show you which agents need attention. Review diffs, browse files, search your codebase, and let agents spawn new agents through the built-in MCP server.
Baton
Hi,
I built this because running multiple Claude Code agents across multiple IDE and terminal windows was getting messy, it was all changing quite fast, and nothing out there handled it the way I wanted. I needed one place to see all my agents, review their changes, and spin up new ones without constantly switching between windows.
I've been building Baton from within Baton for a while now, which has been a pretty fun loop. Would love to hear what you think!
Features.Vote
the git-isolated workspace per agent is the detail that makes parallel agents actually practical. without it you're just tab-managing concurrent terminals that will eventually collide on the same files, and the mental overhead of tracking what each agent touched defeats the point. the isolation isn't an implementation detail, it's the whole model.
Baton
@gabrielpineda Yep, thats a must nowadays, and its not well integrated into existing IDEs.
Awesome man! What’s the long term vision for Baton like do you see it becoming an “OS for AI developers” or more like a power tool within existing workflows?
Baton
@lak7 Thanks, more like a power tool for existing workflows.
I think the traditional IDE will increasingly become a system for managing agent developers. A big part of that is making it easy to run more agents in parallel while still giving the human developer clear oversight of what’s happening.
I think there are really two layers to this. One is expanding how many agents you can comfortably run at the same time, which is what Baton is trying to do. The other is that, at some point, the real bottleneck becomes the developer having to manually direct each agent and constantly write prompts for them. That doesn’t scale.
My guess is that the next step could be moving toward higher-level supervisor agents that sit above the individual workers. Instead of managing every agent directly, the developer works more through systems that condense the important information, suggest strategy, and help direct the overall flow of work. I think Claude and OpenAI will probably build this themselves, but I cant see it happening before coding agents are so robust that you dont need humans to review the code anymore.
@tordrt Interesting pov, so basically there will be a pyramid of agents and the developer will sit at the top. Interesting none the less! Will give it a try and let you know my feedback!!
Baton
@lak7 Thanks! Yeah, at least near term future. Im sure at some point the developer will not be needed anymore🥲
@tordrt 🙃