Replit SEO
Replit gives every public deployment an SEO rating, but it comes from a Lighthouse audit, which renders JavaScript in a browser, so an app whose text only appears after scripts run can rate well and still look empty to AI search bots. Scan the published URL without JavaScript to see what those bots get.
Replit’s SEO rating, and what sits outside it
After each successful publish, Replit runs a Lighthouse audit against the URLs in your deployment and shows a rating: Healthy, Needs Work or Weak, from four groups (AI Readiness, Crawlability & Discovery, Landing Page Rendering & Metadata, Performance Proxies). The badge shows the lowest rating across the crawled URLs, not an average, so one weak page sets it. Autoscale, Reserved VM and Static deployments get a rating; scheduled and private ones don’t, since they serve no public pages.
Replit’s own advice for content-heavy sites is a Static deployment, which serves pre-rendered HTML. An app with a backend runs on Autoscale or a Reserved VM, and its pages are whatever that server sends. If the front end renders in the browser, that is an empty shell.
Fecto’s checks complement the rating rather than repeat it. They read the HTML without running scripts, check robots.txt for each AI search bot by name, and look at the response headers, which is where a stray noindex hides.
Common causes, and the fix for each
A client-rendered front end on Autoscale
Fecto check read.main-text-needs-js · Main text is in the HTML
If the server sends an empty root element and the browser builds the page, Lighthouse sees the finished page and AI search bots see nothing. Pre-rendering the public pages at build time, or splitting the marketing pages into a Static deployment, puts the text back.
No sitemap
Fecto check index.sitemap · A sitemap exists
Replit’s guide suggests having Agent generate one. Ask for it to be built from the app’s routes so it stays current, then check /sitemap.xml on the published domain and that robots.txt points to it.
No robots.txt
Fecto check access.no-robots-txt · robots.txt exists
Same guide, same fix. Make sure the new file doesn’t disallow the AI search bots by accident; templates that only allow Googlebot are common.
Agent-written titles that repeat across pages
Fecto check index.meta-tags · Title and description
Replit suggests a unique title under 60 characters and a description around 150 to 160 per page. The Fecto check also wants a canonical, which most generated apps lack, so search engines don’t pick the replit.app address over your custom domain. Your agent writes each description from what that page actually says, not from a keyword list.
Let your agent do it
If your Repl is connected to GitHub, a local coding agent can do the fixes as pull requests and you pull them into Replit. If you only use Replit Agent, your coding agent can still run the scan and turn each failing check into a specific request you paste into Agent, then re-check the published URL.
Replit recommends a custom domain over the replit.app address for trust with visitors and search engines. Connecting it is yours to do in Replit.
No coding agent yet? Install the Claude app, sign in, then come back and choose Claude app.
Set it up by hand
curl -fsSL https://fecto.ai/install | shThen tell it
Use Fecto to check and fix my site.Don’t treat a Healthy badge as proof AI can read you
Lighthouse renders scripts and runs from Replit’s side. It won’t show a firewall that turns AI search bots away or text that exists only after JavaScript. Use the badge for what it measures and check the rest separately.
What it costs
You pay only when the result shows up on your site: +20% more visits or enquiries is $30, +50% is $60, double is $100, and anything below +20% costs $0. Never more than $100 a site a month.
Questions
Can Replit Agent do SEO?
Replit’s guide says Agent can add or rewrite titles and generate robots.txt and a sitemap. It can’t see what AI search bots get from the deployed app, so pair it with a check that reads the published URL.
Does a Static deployment fix SEO?
For pages that need no server, yes: Replit serves them as pre-rendered HTML. An app with a backend stays on Autoscale or a Reserved VM; its marketing pages can live in a separate Static deployment.
Why is my rating Weak when the home page is fine?
The badge shows the lowest rating across every URL Replit crawled, so one weak page sets it. Open the rating’s details to find which URL, then fix that page.
Do private deployments get indexed?
Private deployments serve no public pages, so Replit doesn’t rate them and crawlers can’t reach them. A page that should be found has to be on a public deployment.
Related guides
Checked against Replit docs: improve your app’s SEO and Replit docs: SEO rating on 18 November 2026.