Remote actions & SSH

Two different ways to reach into a server — a safe, whitelisted action, or a full terminal when you need it.

Remote actions

A small, fixed set of actions the agent is allowed to run — never arbitrary shell commands from the dashboard:

Reboot the server
Restart a service (systemd on Linux, a named Windows service)
Kill a process by name
Tail a log file — returns the last N lines of a specific, allow-listed path
Create an SSH user on the box

Requesting an action queues it for the agent, which picks it up on its next poll (typically within 15 seconds) and reports the result back. You'll see the request move through pending → sent → completed/failed on the server's Actions tab.

Web SSH terminal

A real terminal to your server, in the browser — no separate SSH client needed. Requires SSH credentials to be configured for the server (see Getting started for agent vs. SSH mode). Every keystroke sent is logged as a best-effort command audit trail, viewable later from the server's activity log — useful when a team shares access to a box and needs to know who ran what.

Pro/Business feature.

SFTP file browser

A WinSCP-style file browser and transfer tool over the same SSH credentials — browse, upload, download, and edit small text files (configs, scripts, logs) directly in the browser, without opening a separate SFTP client.

Credential storage

SSH private keys are encrypted at rest and only ever decrypted server-side, in memory, for the duration of a single connection — the plaintext key is never sent to your browser and never logged.

Team permissions: terminal and remote-action access to a server are separately grantable per team member — being on a team doesn't automatically mean full shell access to every server. See Teams & permissions.