How To Deploy Your AI Files

Conceptual guides for CMS, hosting platforms, and frameworks. Uploading these files publicly is required for real discoverability.

Getting Started

Quick Start: what to upload and where

Download all 4 files from the Dashboard and upload them to your domain root.

  1. From Dashboard open your site card and download: llms.txt, ai-profile.json, knowledge.json, services.json.
  2. Upload llms.txt at your public root: https://your-domain.com/llms.txt
  3. Upload the 3 JSON files to a stable public path (recommended: root or /ai/).
  4. Verify that each URL returns HTTP 200 and correct content in the browser.
  5. In the Dashboard, use the Check Files button on your site card — it will show which files are accessible and flag any that are missing.
URLs must be permanent. Changing the path causes AI crawlers to lose previously indexed context.

Frameworks

Next.js on Vercel

Publish files in /public so they are served directly in production.

  1. Copy files into /public (e.g. public/llms.txt and public/ai-profile.json).
  2. Commit and deploy to Vercel — files are served automatically.
  3. Check: /llms.txt, /ai-profile.json, /knowledge.json, /services.json.
  4. For schemas and meta tags in Next.js, use the metadata export in each page.tsx or the root layout.

Frameworks

Astro / Netlify / Cloudflare Pages

Use your project static folder and redeploy.

  1. Put the 4 files in public/ (Astro) or your equivalent static assets directory.
  2. Redeploy in Netlify or Cloudflare Pages.
  3. Validate files are public and cache-friendly (no authentication required).

CMS

WordPress

Upload files via SFTP/File Manager to the public root.

  1. Access /public_html (or your vhost root) via cPanel File Manager or SFTP (FileZilla, Cyberduck).
  2. Upload llms.txt and JSON files to the root directory or /ai/.
  3. For JSON-LD schemas in WordPress: use the Yoast SEO plugin (Schema tab) or add a custom HTML block in the <head> via Appearance → Theme Editor → header.php.
  4. If you use caching (WP Rocket, W3TC) or a CDN, purge it after uploading changes.

Tools & Links

Yoast SEO — Schema in WordPress

CMS

Shopify

Shopify root file access is limited; use a technical subdomain or app proxy.

  1. Recommended: serve files from a controlled subdomain (e.g. ai.your-domain.com) using basic hosting.
  2. Alternative: Shopify App Proxy that responds to /llms.txt and JSON paths dynamically.
  3. For schemas in Shopify: edit theme.liquid and add <script type="application/ld+json"> blocks inside the <head>.
  4. Keep URLs stable so crawlers can index them consistently.

CMS

Webflow

For llms.txt/JSON use auxiliary hosting or a technical subdomain.

  1. Webflow does not allow publishing arbitrary files in the root. Use a technical subdomain (e.g. ai.your-domain.com).
  2. Publish all 4 files on that auxiliary host (a simple Netlify/Vercel repo works fine).
  3. For schemas and meta tags in Webflow: go to Page Settings → Custom Code → <head> and paste the JSON-LD and meta blocks.

Server

Self-hosted Nginx / Apache

Copy files to document root and validate permissions and content-type.

  1. Copy files to your document root (e.g. /var/www/html/).
  2. Set correct permissions: chmod 644 llms.txt ai-profile.json knowledge.json services.json
  3. Verify content-type: text/plain for llms.txt and application/json for JSON files.
  4. In Nginx: if needed, force the type in the server block with the types {} directive.

Schemas & Meta

View your current JSON-LD schemas

Before implementing improvements, identify which schemas you already have and on which pages.

  1. Open any page of your website in Chrome/Firefox and press F12 to open DevTools.
  2. Go to the Elements tab and search (Ctrl+F) for: <script type="application/ld+json">
  3. The content of that tag is your JSON-LD schema. There may be several per page.
  4. Quick alternative: in the browser address bar type view-source:https://your-domain.com and search for "ld+json".
  5. Copy the JSON and paste it into Schema Markup Validator to check for errors.
  6. To see the impact on search results, use Google's Rich Results Test.

Tools & Links

Schema Markup Validator (Schema.org)Rich Results Test — GoogleSchema.org — available types

Schemas & Meta

View your current meta tags (og:, twitter:, title…)

Check which meta tags are in your <head> and how social networks and search engines read them.

  1. Open DevTools (F12), go to the Elements tab, and locate the <head> block.
  2. Look for: <meta name="description">, <meta property="og:title">, <meta name="twitter:card">, etc.
  3. Quick alternative: in the DevTools Console type: document.querySelectorAll("meta") to list all meta tags.
  4. To preview how your link looks on social networks, use Facebook Sharing Debugger for og: tags and Twitter Card Validator for twitter: tags.
  5. For a full visual analysis, paste your URL into metatags.io — it shows preview cards for Google, Facebook, Twitter, and LinkedIn.

Tools & Links

metatags.io — full visual previewFacebook Sharing Debugger (og:)Twitter Card ValidatorGoogle Search Console — URL Inspection

Schemas & Meta

Implement JSON-LD schema suggestions

Add or improve structured markup based on what LLMsArchive recommends in the Improvement Plan.

  1. Open the Dashboard and go to Schema & SEO Meta Suggestions for your site.
  2. For each suggestion marked + new, copy the recommended JSON-LD.
  3. Paste it inside a <script type="application/ld+json"> tag just before </body> on the indicated page (it also works in the <head>, but placing it in the body is the recommended location).
  4. For suggestions marked ~ improve, find the existing schema (see guide above) and update the indicated fields.
  5. Save and deploy. Then validate with Schema Markup Validator and Rich Results Test.
  6. Changes will be reflected in the next scheduled visibility check. Keep in mind that LLMs may take some time to index the new information.
Always implement schemas on the exact pages indicated under "implement on:". A Product schema on the homepage has far less impact than on the actual product page.

Tools & Links

Schema Markup ValidatorRich Results Test

Schemas & Meta

Implement meta tag suggestions (og:, twitter:, description…)

Add or update the meta tags LLMsArchive recommends to improve SEO and social sharing.

  1. In the Dashboard, go to Schema & SEO Meta Suggestions → Meta Tag Suggestions section.
  2. Each suggestion includes the full HTML code ready to copy (e.g. <meta property="og:title" content="…">).
  3. Paste those lines inside the <head> of the indicated pages. If the meta already exists, replace the content attribute value.
  4. meta description: one per page, max ~160 characters, descriptive with the main keyword.
  5. og:image: point to an image of at least 1200×630 px for proper display on social networks.
  6. twitter:card: use "summary_large_image" for most pages.
  7. Verify the result with metatags.io before and after applying.

Tools & Links

metatags.io — visual previewOpen Graph Protocol — reference

Validation

Full validation checklist

Check AI files, schemas, and meta tags before considering any improvement complete.

  1. ✓ AI files: all URLs return 200 and are not blocked by robots/firewall/auth.
  2. ✓ Content is up to date: files match the latest published version in LLMsArchive.
  3. ✓ Schemas: no errors in Schema Markup Validator and Rich Results Test.
  4. ✓ og:/twitter:: meta preview correct in metatags.io.
  5. ✓ Google Search Console: no coverage errors after implementing changes.
  6. ✓ Changes will be visible in the next scheduled visibility check in LLMsArchive. Keep in mind that LLMs may take time to index the new information.

Tools & Links

Schema Markup ValidatorRich Results Test — Googlemetatags.io — og: / twitter:Google Search Console
How To Deploy Your AI Files | LLMsArchive