Ship and scale your application to production without devops overhead, right from your IDE.
import type { Metadata } from 'next'import { Geist, Geist_Mono } from 'next/font/google'import { ThemeProvider } from '@/providers/theme'import { QueryProvider } from '@/providers/query'import { Toaster } from '@/components/ui/sonner'import './globals.css' const geist = Geist({ subsets: ['latin'] })const mono = Geist_Mono({ subsets: ['latin'] }) export const metadata: Metadata = { title: 'dFlow', description: 'Ship from your IDE.',} export default function RootLayout({ children }) { return ( <html lang='en'> <body> <ThemeProvider>{children}</ThemeProvider> </body> </html> )}deploy my ecommerce site on AWS EC2 using dflow MCPdflow MCP
dFlow connects all three bring your own infrastructure, ship any App, Database, Docker Image and drive deployments from the coding agents you already use.
Your cloud
+ bare metal · VPS · on-prem
Your stack
+ any Docker image · 100+ templates
Driven by your AI
dFlow plugs straight into the AI editor you already use. Turns your editor into a deploy console. Go from prompt to production in few minutes.
{ "mcpServers": { "dflow": { "url": "https://app.dflow.sh/api/mcp" } }}Add the dFlow MCP in Cursor, Claude, Codex, or any MCP-compatible editor. One config and your agent gets the full dFlow toolset in chat: apps, services, envs, and deploys.
convert my blog website into a dFlow template by attaching a MongoDB database using dflow mcp
Ask the agent to convert the working tree into a dFlow template, dFlow infers your stack, services, and dependencies and outputs a reproducible template no hand-written YAML.
Tell the agent where to run: Hetzner, EC2, a managed worker, and more. dFlow provisions, builds, and rolls out; you get live deploy logs in chat the moment it goes live.
Ship products and client projects faster, avoid surprise PaaS bills, and keep full ownership from first deploy to reliable scale.
The new DevOps loop
When a deployed service crashes, stream deployment logs straight into your editor, resolve with your favorite AI tools, apply the fix, and push back to production without leaving your workflow.
Deployment timeline
One bad deploy or a runaway loop can turn into a four-figure cloud invoice. dFlow bills what you planned: fixed compute you provision.
INV-456789
$8,342.57
Due in 15 days
Tailscale-ready mesh networking: connect services across regions without exposing ports.
Deploy public or private Git apps and Docker images, plus the databases you rely on: Redis, MongoDB, MySQL, Postgres, and MariaDB.
Organization roles with granular RBAC so the right people can deploy, configure, or view without sharing full admin access.
You own your infrastructure from day one. Everything deploys on your machines, with full control over your data and privacy.
Bring your own cloud
+ many moreStart self-hosted for free, keep full control of your stack, and add managed worker nodes only when your growth needs hands-off scale.
Quick answers about self-hosting, managed worker nodes, pricing, and what to expect as you scale with dFlow.