AgentsRoom Activity tab : every command your agent runs on your server, live
Filmed in the real app: an agent is asked to check free disk space on a staging server, run a health check script and tail the nginx log. The three ssh_exec calls land one by one in the Activity tab of the right panel, each with its output, its exit code and its duration, then the journal of that server opens in full and a button leads back to the Servers list with the search already filled in. The journal never leaves the computer.
Read the full feature pageTranscript
Every command your agent runs on your server, live on your screen. This is AgentsRoom, the desktop app where your AI coding agents work next to your saved servers and databases. A new tab sits next to Files, Changes and Tests.
It stays empty until an agent touches one of your machines. So I ask one to look at a staging box: free disk space, a health check script, then the last lines of the nginx log. It connects, and the server appears straight away, with the command it is running.
They land one at a time: the command, the agent that ran it, and a green dot while it is still going. A chip above the composer tells you which machine is busy. And the agent's own header carries a marker for every server it has touched.
Three calls later, each one shows its exit code and how long it took. One click opens the journal of that server: every call in a row, with its output, the way a terminal shows it. The health check, service by service.
The log tail, back in under a second, with its exit code. From there, one button takes you to that machine in your Servers list, search already filled in. The journal never leaves your computer.
The Activity tab is in AgentsRoom, and the link is in the description.