Fix SEO on a vibe-coded Lovable, Bolt or Cursor site

    Updated

    How to use it

    Copy the prompt, replace the placeholders with your own details, and paste it into ChatGPT, Claude or Gemini.

    Save and favourite prompts with a free account.

    The Site Loaded Perfectly. View Source Was Almost Empty.
    Read the story behind this prompt

    The Site Loaded Perfectly. View Source Was Almost Empty.

    Read the article
    Tap any variable to replace it with your own value:

    AI prompt

    You are a technical SEO engineer who specialises in sites built with AI coding tools such as Lovable, Bolt, Cursor, Claude Code, v0 and similar. These sites are usually single-page apps, so the failures differ from a typical WordPress or Shopify audit.
    
    Task: audit [SITE URL] and, if I provide it, the [CODEBASE OR REPO]. Then implement or hand back the exact fixes needed so search engines can discover, render and index the important routes.
    
    Inputs:
    - [SITE URL]: the live site
    - [CODEBASE OR REPO]: source I can paste or point you at, or "live site only"
    - [BUILDER]: Lovable, Bolt, Cursor, Claude Code, v0, or other
    - [FRAMEWORK]: if known (Vite and React, Next.js, Remix, TanStack Start, or unknown)
    - [ROUTES THAT SHOULD RANK]: the pages that matter, for example home, pricing, blog, docs
    - [BRAND AND DEFAULT TITLE]: how the site should appear in search and on social shares
    - [SEARCH CONSOLE ACCESS]: yes or no, and whether the property is already verified
    
    Work in this order. Inspect the existing codebase or the live HTML before changing anything. Do not break working features.
    
    1. Rendering and crawlability. Check whether important content, titles and links exist in the first HTML response, or only after JavaScript runs. Note the hosting setup (static host, server rendering, prerender, or client-only). If the app is client-rendered, propose a concrete prerender or server-rendering path that fits [BUILDER] and [FRAMEWORK], including which routes to prerender.
    
    2. Per-route titles and meta. Confirm every public route sets a unique title, meta description and canonical in the document head. Watch for a single shell title reused across the whole app. Write the missing titles and descriptions for [ROUTES THAT SHOULD RANK].
    
    3. Open Graph and social tags. Check og:title, og:description, og:image, og:url and twitter:card on each public route. Flag missing or site-wide identical tags, and specify the image file to use.
    
    4. robots.txt and sitemap.xml. Check both files exist, are served with the right content type, list only canonical 200 URLs, and that robots.txt points at the sitemap. Write complete files for [ROUTES THAT SHOULD RANK], and say where to put them in this repo (usually the public folder).
    
    5. Canonicals, trailing slashes and host variants. Confirm one canonical host and a matching canonical tag on each route. Flag duplicate paths such as index.html, hash routes, or trailing-slash copies.
    
    6. JSON-LD. Check for Organization and WebSite schema in the served HTML head, plus any page-level types that fit. Write valid JSON-LD. Do not invent ratings, review counts or prices.
    
    7. Crawlable links and 404 handling. Confirm primary navigation uses real href links. Click-only buttons without an href will hide those routes from a basic crawl. Confirm unknown paths return a real 404 status, rather than a 200 with a client-side "not found" screen. Give the exact router or host change needed.
    
    8. Indexation hygiene. Search the repo and rendered HTML for noindex, nofollow, X-Robots-Tag, and assets blocked from crawlers. Only remove noindex when it is clearly unintentional.
    
    9. Search Console. If [SEARCH CONSOLE ACCESS] is no, give the exact verification steps and how to submit the sitemap. If yes, say what to inspect first: coverage, why pages are not indexed, and the rendered HTML on the URL Inspection tool.
    
    10. Verify each change. After edits, list how to confirm them: curl the HTML for titles and JSON-LD, open /sitemap.xml and /robots.txt, check status codes on key routes and on a made-up path, and run the Rich Results test.
    
    Constraints:
    - Prefer fixes in this project's existing head, router and static-file setup. Do not add a new framework just to ship tags.
    - Mark anything you could not verify, and anything that still needs a manual step such as DNS, Search Console, or prerender hosting.
    - If you later need a broader technical audit, use https://prompt-king.ai/prompt/technical-seo-audit-checklist. For a mobile-first pass, use https://prompt-king.ai/prompt/mobile-first-seo-optimization-checklist.
    
    Output:
    - A findings table with area, what you found, evidence (file, route or HTML), fix, owner (you can patch, or I must do this), and priority.
    - The patched files or ready-to-paste snippets: robots.txt, sitemap.xml, head tags, JSON-LD, 404 config, and prerender or server-render config.
    - A short "still manual" list covering Search Console, DNS, image assets and hosting.

    Share this prompt