From 1 Agent to 30
Lessons from scaling an AI agent fleet.
1-3 agents: Manual management works
You can manage a few sessions with tmux directly. But you'll lose track of what each one is doing.
5-10 agents: You need a dashboard
At this point, you can't keep everything in your head. amux's dashboard shows live status, token spend, and current task for every agent.
10-30 agents: You need coordination
Without a task board and atomic claiming, agents duplicate work. amux's kanban board with SQLite CAS ensures no wasted effort.
Get started with amux
Run dozens of Claude Code agents in parallel. Python 3 + tmux. Open source.
git clone https://github.com/mixpeek/amux && cd amux && ./install.sh
amux register myproject --dir ~/Dev/myproject --yolo
amux start myproject
amux serve # → https://localhost:8822View on GitHub