Monitor every machine via SSH — or install a lightweight agent for even easier setup. Rules engine with webhooks, MQTT, and command actuation. Talk to your fleet through your OpenClaw agent. Process management, GPU orchestration, and predictive alerts — all from one tool.
Define rules in YAML. When a shared machine goes idle with a game running, Leassh kills the process and starts Ollama. When disk fills past 90%, fire a webhook. Nine built-in triggers ship active by default.
SSH into every machine — or deploy agents for zero-configuration monitoring. CPU, GPU, RAM, disk, processes, idle time, temperatures. Windows, macOS, Linux.
Webhooks with template variables, MQTT for Home Assistant, SSH commands on any node. Kill processes, restart services, reclaim GPUs — automatically or on demand.
Rate-of-change analysis on disk, RAM, and temperature. "Disk losing 12 GB/h, critical in ~40 minutes." Not reactive — predictive.
Ask your AI agent "how's the fleet?" and get real-time answers. Proactive alerts in chat. The free plugin monitors — Leassh Pro acts.
Docker or binary: docker run -p 8383:8383 leassh/leassh
Via agent: Install the agent on each machine, enter your pairing code — it connects automatically through firewalls and NAT.
Via SSH: Add machines to fleet.yaml with SSH credentials. No software installed on targets.
Set up triggers, actuation, and limits in YAML or the UI. Fleet dashboard at :8383/fleet.
Choose the plan that fits your infrastructure.
| Feature | Essential | Family | Pro |
|---|---|---|---|
| Machines | 3 | 10 | Unlimited |
| SSH probing | ✓ | ✓ | ✓ |
| Daily summaries | ✓ | ✓ | ✓ |
| Basic alerts | ✓ | ✓ | ✓ |
| AI screenshot analysis | — | ✓ | ✓ |
| Weekly reports | — | ✓ | ✓ |
| Time limits + enforcement | — | ✓ | ✓ |
| Telegram notifications | — | ✓ | ✓ |
| Rules engine + YAML | — | — | ✓ |
| Webhooks + MQTT | — | — | ✓ |
| SSH actuation | — | — | ✓ |
| Fleet dashboard | — | — | ✓ |
| OpenClaw plugin | — | — | ✓ |
| Self-hosting | — | — | ✓ |
| Full REST API | — | — | ✓ |
| 14-day free trial | — | ✓ | ✓ |
You have two options. The recommended approach is to install the tiny Leassh agent (under 10MB) — just enter a pairing code and it connects automatically. Alternatively, power users can use SSH probing which requires no software on target machines but needs SSH access configured. Both give you the same monitoring data.
Define rules in YAML with conditions and actions. Conditions can reference any probe metric (CPU, GPU, disk, temperature, processes, idle time). Actions include webhooks, MQTT messages, and SSH commands. Nine built-in rules ship active by default — you can customize or disable them.
openclaw-fleet is a free, open-source plugin that connects your Leassh fleet data to any OpenClaw-compatible AI agent. Ask "how's the fleet?" in natural language and get real-time status, proactive alerts, and suggested actions. Available at github.com/leassh/openclaw-fleet.
Yes. The Pro plan includes full self-hosting support. Run docker run -p 8383:8383 leassh/leassh or download the binary. Your dashboard, your data, your infrastructure. No cloud dependency.
CPU, GPU, RAM, disk usage, temperatures, running processes, active users, idle time, and network interfaces. All data is collected via read-only SSH commands. Leassh never modifies your machines unless you define actuation rules that explicitly do so.
Monitoring is read-only: Leassh probes metrics via SSH. Actuation means Leassh can also execute commands — like killing a process, restarting a service, or running a script. Actuation is opt-in and requires explicit rules in your YAML config. Only available on the Pro plan.
Yes. If you already have SSH access, you can install the agent without physically touching the machine. Run the install script via SSH with your pairing code. On macOS, the Screen Recording permission still needs one-time physical approval — everything else works immediately.
Rules, webhooks, and SSH actuation. Deploy in under three minutes.