Free tool

I built this with AI. Is it actually safe to launch?

Paste a public URL. In a few seconds you get a Launch Score out of 100 and a short list of what would embarrass you in front of your first customers — the same issues that ship quietly when an app is built faster than it can be verified.

Security postureHTTPS, security headers, cookie flags, and whether .env or .git files answer publicly.

Broken thingsDead links on the landing page, missing robots.txt and sitemap.xml.

The basicsTitle, meta description, mobile viewport, favicon — the things search and sharing depend on.

First impressionsTime-to-first-byte and the weight of the HTML document.

How the score works

Four categories, weighted by how much they hurt.

Security counts for 40 of the 100 points, reliability 25, the basics 20, and performance 15. Every check reports what it actually measured — a probe that could not run is reported as skipped, never guessed. A perfect score here means the public surface of your app is in order; it is not a substitute for reviewing code, database rules, or payment handling.

One honest caveat: some sites try to block automated visitors, and where a site serves our scanner a challenge page instead of the real one, the report says so — those results are marked skipped rather than guessed. The desktop app already coordinates the AI agents that build the software; repo-level checks — the ones that find tenant leaks and unhandled payment states — are what we are building toward next.