HN Explorer

Show HN: Amux – A tmux-based multiplexer for running parallel Claude Code agents (amux.io)

2 points by Beefin ↗ HN
Hi HN, I'm Ethan. I built amux because I was spending more time switching between terminal tabs than actually building things.

amux is an open-source agent multiplexer that lets you run, monitor, and control multiple headless Claude Code sessions from a single dashboard — in your browser, on your phone, or from the terminal.

The problem: I run 5-10 Claude Code agents at a time across different repos. Keeping track of which one is waiting for input, which one is working, and which one broke something was chaos. I needed a control tower.

What it does:

- Spin up any number of Claude Code sessions, each isolated in its own tmux pane

- Live status detection (working / needs input / idle) via SSE — know at a glance which agents need you

- Web dashboard installable as a PWA on your phone. Monitor and send commands from the couch

- Multi-pane grid view to watch multiple agents side by side

- File attachments — paste images, drag-and-drop files directly to agents

- Built-in kanban board for tracking work across all your agents

- Token usage stats so you know what each agent is costing you

- Tailscale integration for secure remote access with zero config

The entire thing is a single Python file with zero dependencies beyond Python 3 and tmux. No build step, no npm install, no Docker. Clone, run `install.sh`, done.

I use this every day to coordinate agents working on different microservices simultaneously. The phone PWA is surprisingly useful — I'll kick off a batch of tasks and check in from my phone while doing other things.

Everything is also exposed via REST API so you can script orchestration workflows with curl.

MIT licensed: https://github.com/mixpeek/amux

Site: https://amux.io

Happy to answer any questions about the architecture or how I use it day to day.

5 comments

[ 4.4 ms ] story [ 24.1 ms ] thread
tumf
Awesome project — I’ve wanted exactly this for coordinating headless agent sessions across repos.
Beefin
glad you like it, feel free to suggest anything. i've been using this to be productive all day on my phone.
AlexCalderAI
Really nice approach. Running parallel agents is the obvious next step after single-agent works, but the coordination + cost tracking becomes the blocker.

One thing I'd add: make cost-per-agent visibility mandatory from day one. When you've got 5-10 agents running, token burn compounds fast. Each agent needs its own cost dashboard.

We bake this into every agent setup now—cost awareness prevents expensive mistakes.

Beefin
One thing we’ve been thinking about with Amux is that the unit of compute shouldn’t just be the terminal session—it should be the agent itself. That means each pane/session can expose things like:

* tokens in / tokens out * cumulative run cost * model + pricing tier * runtime duration * optional budget caps

So when you spin up 5–10 agents, you can immediately see which one is burning tokens or looping.

Longer term I’d love for Amux to treat agents a bit like processes in `htop` where you can see resource usage across all agents in one place and kill/restart the expensive ones quickly.

Curious how you're currently surfacing cost in your setups — logs, dashboards, or something inline with the agent runtime?