Download Lesuto Hub & Seller - Shop and sell on iOS and Android
Integrations
Chameleon Render Engine
Give search engines and social apps real HTML on Commerce SDK sites, with titles and share images.
For merchants
DeploymentMerchantFree
HTML For Crawlers
Search engines and social apps get a finished HTML page with titles, descriptions, and share images.
Shoppers Stay Live
Visitors keep the interactive store. Cart and checkout never wait on a snapshot.
Cached And Fresh
Pages are cached so crawlers get a fast response. Catalog changes invalidate the cache.
Features
- 24-hour render cache
- Automatic meta tag generation (title, OG, Twitter)
- Concurrent render limiting
- Domain-based access control
- Finished HTML for search engines and social previews
- Health monitoring dashboard
- JSON-LD structured data injection
Requirements
- Active Commerce SDK deployment
- Bot detection configured on your web server
How it works
Trace a visitor from the URL to HTML
Pick a shopper, Googlebot, Facebook, GPTBot, or an audit crawler. The path matches production: edge detection, api.lesuto.com/api/v2/prerender, 24-hour cache, and origin fallback.
Live trace
Follow one request
Search and social crawlers on catalog paths. Included on Lesuto-hosted stores.
Edge cases
Googlebot on Cloudflare / Lesuto-hosted
- Visitor
- Your site
- Bot check
- Lesuto API
- Cache
- Headless render
- HTML snapshot
In flight
Googlebot requests /products/teak-lounge
User-Agent: Googlebot/2.1
Request log
01 info
Googlebot requests /products/teak-lounge
User-Agent: Googlebot/2.1
Lesuto-hosted stores already route matching bots. On your own host, add a Cloudflare Worker or use the WordPress plugin. Confirm with curl before you cut DNS.
The problem
Crawlers do not shop like people
Google can render JavaScript later, slowly. Facebook, LinkedIn, Slack, iMessage, and most AI crawlers fetch source HTML only. If that source is an empty JavaScript shell, the page is invisible.
Without Render Engine
- ✗Search engines see an empty page first
- ✗Social shares show no image or title
- ✗Chat unfurls fail in Slack, iMessage, WhatsApp
- ✗AI crawlers may never see your copy
With Render Engine
- ✓Matching crawlers receive full HTML
- ✓Open Graph and Twitter cards work
- ✓Link previews work in social and chat apps
- ✓Visitors still get the interactive site
Who it serves
Search, social, chat, and AI, with honest routing
One snapshot path. Different hosts decide which user-agents are forwarded.
Search engines
Google, Bing, DuckDuckGo, Yandex, and Baidu get the rendered HTML, not the empty shell.
Social networks
Facebook, Instagram, LinkedIn, X, and Pinterest unfurls can show title, image, and price.
Chat apps
Slack, iMessage, WhatsApp, Discord, and Telegram can preview a URL instead of a blank link.
AI crawlers
The WordPress plugin snapshots common AI bots. Cloudflare does not, unless you extend that Worker. Audit bots are never prerendered.
Setup
Three steps to crawlable JavaScript pages
Connect a domain, route crawlers, confirm with curl. Commerce SDK and WordPress are first-party hosts. Any JavaScript frontend on that domain can use the same path.
Connect your domain
A site key ties the Render Engine to hostnames you own. Visitors still get your live JavaScript app.
Route crawlers to the Render Engine
On a Lesuto-hosted store this is already on. On your own host, add a Cloudflare Worker, Apache or nginx rule, or connect the WordPress plugin.
Read the install guideVerify before you share
curl with a Facebook or Googlebot user-agent should return 200 and X-Prerender. Then confirm Google Rich Results, Facebook Sharing Debugger, and LinkedIn Post Inspector.
Questions
How routing actually behaves
The visualization above is the same decision tree the API, Cloudflare Worker, and WordPress plugin use.
Is this only for the Commerce SDK?
No. Any JavaScript frontend has the same crawler problem. Connect a site key for your domain and route bots. Commerce SDK and WordPress are first-party hosts. SDK pages wait for a ready event so catalog HTML is complete. Other JavaScript on that domain snapshots after the render budget.
Do visitors ever get the prerendered page?
No. The Render Engine inspects the user-agent and only snapshots matching crawlers. Visitors keep the live JavaScript site, including cart and checkout when those are on the page.
Which crawlers get HTML snapshots?
Search engines (Google, Bing, DuckDuckGo, Yandex, Baidu) and social unfurlers (Facebook, LinkedIn, X, WhatsApp, Slack, Discord, Pinterest, Telegram). The WordPress plugin also routes common AI crawlers such as GPTBot and ClaudeBot. Cloudflare and Lesuto-hosted routing stay on the search plus social list unless you change that Worker on purpose.
Why are ChatGPT and audit bots treated differently?
SEO audit tools (Ahrefs, Semrush, Screaming Frog) are classified in the dashboard and are not prerendered, so they cannot spend render capacity. AI crawlers are classified everywhere. Only the WordPress plugin host hook prerenders a common AI set today.
What happens if the renderer is down or the domain is wrong?
The API returns 403 when the hostname is not on the site key, and 502 or 503 when a render fails. Your Worker or plugin then serves origin HTML so crawlers are not left with a blank error page.
How fresh is the HTML?
Successful snapshots are cached for 24 hours. Catalog changes invalidate the cache. The first crawl for a URL runs a headless browser (12 second budget). Later crawls are served from cache.
Better together
How it fits the rest of Lesuto
The Render Engine is the HTML snapshot layer under any JavaScript frontend you connect, including SDK, WordPress, and Studio pages.