Scale your AI agent fleet across unlimited machines. One unified dashboard to manage agents running across your entire infrastructure—from your laptop to cloud servers.
"Developers in 3 cities. Agents on local machines and our cloud. They coordinate."
Manage agents across unlimited machines from one dashboard
Peers connect via WebSocket and automatically register themselves. No complex configuration—just add the SSH connection and you're done.
Real-time health checks show peer status, CPU usage, memory consumption, and network connectivity. Know immediately when something goes wrong.
All peer sessions appear in one hierarchical view. No switching between dashboards—manage everything from a single interface.
Create, attach, and manage tmux sessions on remote peers just like local sessions. Full terminal streaming over SSH with low latency.
Monitor CPU, memory, disk usage, and active sessions per peer. Identify bottlenecks and optimize workload distribution.
Works on macOS, Linux, and Windows WSL2. Mix and match different platforms in your peer mesh—the dashboard handles them all.
Simple architecture, powerful capabilities
Add SSH connection details in Settings (hostname, port, username). AI Maestro must be installed on the remote machine.
Dashboard establishes WebSocket connection via SSH tunnel. Peer registers itself with hostname, platform, and capabilities.
Peer sends list of all tmux sessions. Sessions appear in your dashboard with peer hostname prefix.
Click any remote session to attach. Terminal I/O streams over WebSocket with full xterm.js rendering.
┌─────────────────────┐
│ Your Laptop │
│ AI Maestro │
│ Dashboard │
└──────┬──────────────┘
│
├──[WebSocket]───► Peer 1 (macOS)
│ 10 agents running
│
├──[WebSocket]───► Peer 2 (Linux)
│ 8 agents running
│
└──[WebSocket]───► Peer 3 (Cloud)
15 agents running
Total: 33 agents across 3 peers
All manageable from one dashboardScale beyond a single machine
Large codebases or compute-heavy tasks can overwhelm a single machine. Distribute agents across multiple peers to parallelize work and speed up development.
Test cross-platform code by running agents on macOS, Linux, and Windows peers simultaneously. Catch platform-specific bugs before they reach production.
Keep sensitive code on secure machines while controlling agents from your laptop. Peers stay behind firewalls—only WebSocket connection needed.
Share a pool of powerful peer machines with your team. Everyone connects to the same peers but manages their own agents independently.
Set up your first peer in 5 minutes
curl -fsSL https://raw.githubusercontent.com/23blocks-OS/ai-maestro/main/scripts/remote-install.sh | sh
# In AI Maestro dashboard Settings page:
# Add new peer:
# - Hostname: remote-peer.example.com
# - Port: 22
# - Username: your-username
# - SSH Key: ~/.ssh/id_rsa (optional)Peer sessions automatically appear in your dashboard. Click any session to attach and start coding!
Start with one peer, scale to hundreds. The dashboard stays the same.