From upload to live URL — in under a minute.

MoonBase takes whatever AI built and gets it running in production. Here’s exactly what happens between upload and live.

Step 1 — Upload your project

Drag a ZIP into MoonBase, or push directly from your build pipeline. Any origin is accepted: Claude, Cursor, Lovable, Bolt, v0, Replit, or a manual export from any tool.

No CLI. No configuration file. No moonbase.json.

Step 2 — The Inspector reads your project

Before anything runs, MoonBase performs a full structural analysis:

• What framework is this?
• Where are the build artifacts?
• Are there missing environment variables?
• Is the output structure what the runtime expects?
• Are there security red flags (hardcoded keys, exposed credentials)?

This happens automatically. You see the result in seconds.

Step 3 — The Compatibility Engine repairs it

Most AI-generated projects have at least one structural issue that would break a standard deployment. MoonBase fixes them:

• Incorrect output paths
• Missing or misconfigured environment references
• Build artifacts placed in unexpected locations
• Framework configuration mismatches

If something can’t be repaired automatically, MoonBase tells you exactly what needs fixing and why.

Step 4 — Deployment Recipe executes

Based on what the Inspector found, MoonBase selects the right deployment path:

• Static — for HTML/CSS/JS projects with no server component
• Dynamic — for server-side rendered applications and API routes
• Hybrid — for frameworks like Next.js that mix both

The recipe runs automatically. No decision required from you.

Step 5 — Your site is live

You get a live HTTPS URL. Every version is preserved. Roll back at any time.

That’s it.

One path for every origin

The same process handles every project, regardless of what generated it or what framework it uses. There’s no “compatible” list. If it runs on Node.js or serves static files, MoonBase can host it.