Approve your remote agents
from your phone.

roam runs autonomous agents on your own servers — detached, budgeted, sandboxed. The moment one reaches for something destructive, it stops and asks. Get the ping, tap approve or deny, from anywhere.

Get early access Try the free CLI

└─ phones home over https ─┐

roam · now

Agent on prod-db-01 wants to run a destructive command

rm -rf /var/lib/postgresql/old

How it works

01 · dispatch

Send it and walk away

roam send --to prod-01 --goal "…" ships the binary to the box and runs the agent detached. One static binary — no Python, no runtime on the server.

02 · it parks

Nothing destructive runs unasked

Hit rm -rf, git push --force, DROP TABLE, shutdown… and the agent freezes and asks. Budgets and a workdir sandbox bound the rest.

03 · you decide

From anywhere, in one tap

Approve, deny (the agent adapts), or stop. See every agent across every box — live status, token cost, the full journal — in one place.

Why not just ssh?

Because you're not always at your laptop.

The free CLI already lets you approve over ssh. But the agent parks at 2am, or you're on café wifi that can't reach the box. The hosted layer is the agent phoning home — so the decision comes to you, wherever you are.

What early access gets you

→ push / email / WhatsApp when an agent parks
→ approve & deny from the web or a tap
→ fleet view across all your servers
→ aggregated token cost + audit trail
→ scheduled dispatches

Early access

Leave agents running. Sleep fine.

The roam CLI is free and open source today. Join the waitlist for the hosted panel — notify, approve from anywhere, and fleet view.

✓ You're on the list — we'll be in touch.