Ship
Deploy any app from GitHub. Paste the link; Arca builds it, runs it and gives it an address.
- GitHub
- Your app
- Postgres
- Cloudflare
Infrastructure for AI agents.
Yours, end to end.
Needs a Linux machine (x86_64 or ARM) with Incus already running. It sets Arca up to look only: Arca changes nothing on the machine it runs on until you say which parts are Arca's.
Or download it yourselfDeploy your own app or pick a ready one, on your machines or ours.
Deploy any app from GitHub. Paste the link; Arca builds it, runs it and gives it an address.
A team of agents that is always on, with their memory and their messages.
Your agents on every channel people already use.
Give your agents a phone of their own.
Give your agents a computer of their own, with a browser and a screen.
Your GitHub builds on your own machines.
A database in one request, with snapshots.
Run open models on your own graphics cards, next to the agents that use them.
The Mac on your desk. The server in the closet. The laptop in a drawer. Add machines as you go; Arca shows them as one place.
Example figures. Arca reads the real ones when it starts.
One person, a laptop and a rented server. The whole stack runs on them, and your agent keeps it running while you build.
Agents, their memory, their tools and their data on machines the team controls. Need another service? Ask for it.
A separate project for every client, each on the machines you choose. Hand over the key when the work is done.
Your own graphics cards, shared between your models and the agents that use them.
The laptop under the counter runs the site, the orders and the agents that help.
Tell your agent what you need. It prepares the change. You approve with a passkey. It runs.
curl -fsSL https://get.automagik.si/arca | sh
Early access is opening. Leave your email; your agent can sign you up too.






