Published on build. Found by everything.
Every publish lands on a global CDN with HTTPS and your own domain — then Kliv maps your site, writes the sitemap, and serves prerendered pages to the crawlers that can't run your app. Humans get the live app; everything else gets the truth in plain HTML.
Hosting & SEO
Not a prerender service bolted on — part of publishing
Modern app builders ship single-page apps that are invisible to anything that doesn't execute JavaScript — social unfurlers, AI crawlers, most search engines that aren't Google. The usual fix is a third-party prerender proxy with its own bill, its own cache, and its own idea of when your site changed. Kliv builds the fix into publishing itself: the same pipeline that deploys your app crawls it, captures each page after it has fully rendered, and serves those snapshots to crawlers — pinned to the exact version your visitors see, refreshed when you publish.
From publish to cited
The distance between "my app is live" and "my app gets found and quoted" is usually a stack of plugins and a consultant. Here it's one pipeline. Follow a publish all the way through.
Publish
You publish; seconds later the site is live on a global CDN with HTTPS. There's nothing to deploy and nothing to configure — publishing is a property of building, not a separate project.
Your domain
Connect your own domain when you're ready. The site, the checkout, the receipts, the sitemap — everything carries your name, not a builder's subdomain.
The spider maps your site
After a publish, Kliv crawls the site the way a search engine would: starting from your routes, following every internal link, rendering each page in a real browser and waiting for the content to settle before reading it.
The sitemap writes itself
Pages that render cleanly and aren't marked noindex become your sitemap.xml, advertised in robots.txt automatically. The lastmod dates move when content actually changes — so crawlers learn that your sitemap tells the truth.
Snapshots are captured
Each page is stored as plain HTML exactly as it rendered — data loaded, content settled, titles and structured data in place — and pinned to the version you published. A new publish supersedes them; crawlers never see a stale mix of old and new.
Crawlers get the real thing
Visitors get the live app. Crawlers that don't run JavaScript — AI assistants' crawlers, social preview bots, non-rendering search engines — get the snapshot: your actual content instead of an empty application shell. When someone asks an AI assistant a question your site answers, there is something real to cite.
Every page is audited
The Search & SEO screen lists every page the spider found: in the sitemap or not, snapshotted or not, its inbound links, and its on-page issues — missing descriptions, overlong titles, skipped headings, images without alt text, thin content. Open any page and compare the live app with exactly what a crawler sees.
What's in the box
One system from publish to search — each piece reading the same crawl of the same site.
The sitemap writes itself
Generated from a real crawl of your published site — not a plugin you keep in sync. Parameterized routes are covered page by page: every product page that's linked gets discovered and listed.
Snapshots of the real thing
Captured after each page has fully rendered, stored as plain HTML, pinned to the published version. Format changes for crawlers; your visitors' app doesn't change at all.
Quotable by AI
ClaudeBot, GPTBot, and PerplexityBot don't run JavaScript. With snapshots they read your actual content — so AI assistants can cite your pages instead of skipping them.
A page-by-page audit
Titles, descriptions, heading structure, alt text, canonical tags, thin content — checked per page with an indexability verdict and the reason behind it, not a score without an explanation.
You decide who reads
Prerendering changes what crawlers receive, never who gets in. Your robots.txt stays yours — including rules that keep AI crawlers out of some or all of your site, if that's your call.
Links that unfurl properly
Slack, iMessage, and social preview bots don't run JavaScript either. Snapshots give them real titles, descriptions, and images to show the moment someone shares your link.
Common questions
Does Google need this?
No — Googlebot executes JavaScript and can index your app without prerendering, and Kliv serves snapshots to Google only if you deliberately enable the top rung. The wins are everywhere else: AI crawlers, social unfurlers, and the many search engines that don't render JavaScript.
What do AI crawlers actually see?
The snapshot: your page's real content in plain HTML, captured after it fully rendered. Without it, a crawler that can't run JavaScript sees an empty application shell. With it, an AI assistant answering a question in your domain has your actual words to quote — and a link to attribute them to.
Do I have to configure anything?
Prerendering is a deliberate ladder in your account settings: off by default, then a capture-only test mode so you can inspect snapshots before anyone else sees them, then serving to non-Google crawlers, then — after an explicit confirmation — everyone. The sitemap has its own switch and can run even with prerendering off.
What ends up in the sitemap?
Pages that render successfully and aren't marked noindex, discovered by actually crawling your published site. Pages behind sign-in aren't included — the spider sees your site as a signed-out visitor sees it, which is exactly what a search engine sees too.
Can I keep crawlers out?
Yes — access control is your robots.txt's job, and it stays yours. Add rules to exclude specific crawlers, including AI crawlers, from some or all of your site; Kliv appends the sitemap advertisement without ever touching your rules.
What does the audit check?
The on-page fundamentals, per page: title and description presence and length, heading structure, canonical tags, image alt text, social tags, thin content, language attributes — each with an indexability verdict and its reason. And for any page, a side-by-side: the live app next to the snapshot a crawler receives.
Publish something worth finding
Your pages carry real titles, descriptions, and structured data because the same system that builds them publishes them. Describe your app — being found comes built in.
