Scan Your Site During Testing

Catch SEO, schema, performance, and content issues before you deploy. Three ways to point Greadme at a site that isn't live yet.

★ Recommended

Use a Real Domain

Buy a domain from any registrar and use it as your dedicated testing URL, kept separate from your real production domain. The most stable, team-shareable setup — perfect for long-term use.

  • Stable URL that never changes
  • Cleanly separated from production
  • Shareable with your whole team
  • No rate limits, no warning pages
Namecheap
GoDaddy
Cloudflare
Free & stable

Use Free Hosting

Deploy your dev build to a free hosting service. You get a permanent, public URL on a vendor subdomain like myapp.netlify.app — no domain to buy, no setup hassle.

  • Completely free
  • Stable URL across sessions
  • Deploy from Git in minutes
  • No domain purchase required
Vercel
Netlify
Render
CF Pages
GitHub Pages
Push your code, get a public URL — scan as often as you want
Quick start

Use a Tunnel

Run ngrok http 3000, get a temporary public URL, paste it into Greadme. Done in under a minute — no deploy, no signup beyond ngrok itself.

  • Set up in under a minute
  • No deploy needed — tunnels to localhost
  • Free tier available
  • Great for fast iteration while you build
ngrok
$ngrok http 3000
Forwarding https://abc123.ngrok-free.app
localhost:3000

Which method should you pick?

MethodSetup timeCostSame URL every timeBest for
Real domain~15 minutes, plus DNS propagationPrice of the domainYesLong-lived staging a whole team scans
Free hostingMinutes, from a Git pushFreeYesSolo projects and side projects
TunnelUnder a minuteFree tierNo — a new URL each restartA quick check while you're still coding

What has to be true before a scan works

Greadme fetches your page from our servers, the way a search engine would, so the address you paste in has to be reachable from the public internet. Three things commonly stop a site that's still in testing:

  1. The URL has to be public

    localhost, 127.0.0.1, private ranges such as 10.x, 172.16–31.x and 192.168.x, and hostnames ending in .local or .internal are refused before any request goes out. That's exactly the gap a tunnel fills: it hands your local server a public address for as long as it runs.

  2. robots.txt has to allow the page

    Testing sites usually ship with Disallow: / to keep search engines away. If the page you enter is disallowed, the crawl stops before it starts — and that attempt is not counted against your quota. Relax the rule for the page you want to test, or point the scan at a copy that permits crawling.

  3. Your firewall has to let us through

    If the site sits behind a WAF (a web application firewall, the filter that blocks suspicious traffic), the scan comes back reporting a block page instead of your content. Allow the scanner with Organization Auth, which proves the request is ours with a secret key rather than a user-agent string anyone can copy.

Expect one “failure” that isn't one

Sites in testing normally carry a noindex tag, so the report will flag Search Engines Blocked from Page. On a testing URL that result is correct — the tag is doing its job. Note it, ignore it there, and confirm it has disappeared when you scan production.