Skip to content
Platform overview

Prompt to production, end to end

One pipeline turns a request into live infrastructure and keeps it operating — from chat, the dashboard, or an agent over MCP.

90+
MCP tools, same checks as the dashboard
microVM
Firecracker-isolated per service
24h
restore window on every deletion
3 × 11
roles × permission categories per org

How it works

Parse, plan, confirm, run

Chat and MCP run the same pipeline. Nothing deploys until you confirm the plan.

  1. 01

    Parse

    Your request becomes a structured intent.

  2. 02

    Plan

    An editable plan card lists every resource it will create.

  3. 03

    Confirm

    Nothing runs until you confirm.

  4. 04

    Run

    It builds and returns a live HTTPS URL.

What you get

Everything to run it and keep it running

Every part of it is managed from the dashboard and over MCP.

  • microVM-isolated services

    A hardened Firecracker microVM per service.
  • Managed databases

    An isolated PostgreSQL database per project.
  • Automatic HTTPS + domains

    Automatic HTTPS on every URL. Custom domains with managed certificates.
  • Operate and recover

    Logs, rollback, roles, and a 24-hour restore window on anything deleted.

Whole-project Compose export is on the roadmap. Your images, data, and credentials are exportable today.

See it run on your own workload.

Join the waitlist and onboard in a small cohort. Bring a repo, a container image, or a catalog app.