ModKor SEO Suite — AI-Ready SEO for Magento 2

1. What it does

ModKor SEO Suite manages everything search engines and AI assistants read about your store — meta titles and descriptions, canonical URLs, robots directives, social share cards, hreflang, a connected schema.org rich-data graph, and the /llms.txt business summary AI crawlers ingest — from one dashboard, with zero JavaScript added to your storefront.

Highlights

• Meta templates + live Google preview — variables, fallbacks, optional segments, length meters.
• Managed canonicals — exactly one per page; self-canonical pagination; tracking params stripped.
• Robots rules — pattern lines with wildcards; search results noindexed by default.
• Open Graph + Twitter Cards — with live product price and a share preview.
• Hreflang — real per-store URLs, locale-derived codes, x-default.
• Rich data @graph — Organization, founder, WebSite, Store locations, FAQ, Product with ratings, breadcrumbs.
• AI Discovery — /llms.txt + /llms-full.txt, assembled live.

2. Requirements

• Magento Open Source / Adobe Commerce 2.4.x
• PHP 8.1 – 8.4 · Hyvä or Luma
• ModKor_Core (ships with the suite)

3. Installation

composer require modkor/module-seo
bin/magento module:enable ModKor_Core ModKor_Seo
bin/magento setup:upgrade
bin/magento setup:di:compile            # production
bin/magento setup:static-content:deploy # production
bin/magento cache:flush

Defaults are conservative: templates only fill empty metas, and the rich-data graph stays silent until you enter your organization — nothing on your storefront changes until you want it to.

4. Quick start (3 minutes)

1. Open ModKor → SEO. The suite is enabled with safe defaults.
2. On Meta Templates, adjust the product title/description templates — the live Google preview below shows the result with a real product from your catalog.
3. On Rich Data, enter your organization (name, description, profiles) — this switches on the sitewide schema graph and llms.txt content.
4. On FAQ & Locations, add your most-asked questions and your stores (if you have physical locations).
5. Save settings — then view source on a product page and open /llms.txt.

5. The dashboard, tab by tab

Meta Templates

Templates for product and category pages with {variables}, fallback chains like {brand|manufacturer}, and optional [bracket] segments that disappear when empty. Force overwrite is off by default, so hand-written metas always win. Hard max lengths truncate at word boundaries.

Canonicals

One managed rel=canonical per page — any canonical set elsewhere is removed first. Products default to their category-free single URL; paginated pages are self-canonical (Google's current guidance); tracking parameters are stripped.

Robots

One rule per line: /checkout* NOINDEX,NOFOLLOW. Wildcards, comments, first match wins — plus the recommended noindex for internal search results.

Social Cards

Open Graph and Twitter Card tags built from your final metas, the product image and live price — with a share-card preview. A fallback image covers pages without their own.

Hreflang

For multi-language stores: alternates resolved from each store view's real URL rewrites, codes derived from store locales with per-store overrides, and an x-default store. The table shows exactly what will be emitted.

Rich Data

The schema.org @graph: your Organization (with sameAs profiles and contact details), an optional founder Person, WebSite, Store nodes for retail locations and FAQPage — plus Product schema (live price, stock, star ratings) and BreadcrumbList on catalog pages. The graph only emits once your organization name is set.

AI Discovery

Serves /llms.txt (concise) and /llms-full.txt (adds the founder bio): business facts, retail locations, your live top-level category tree, FAQ and key pages — assembled fresh on every request.

FAQ & Locations

Inline editors for the questions and retail locations that feed both the JSON-LD graph and llms.txt. Row order is display order; AI assistants quote these answers verbatim, so write real ones.

Per-store-view overrides

Pick a store view in Scope, flip Override on any section for view-specific settings; everything else inherits from Default.

6. Verifying your setup

1. View source on a product page: your templated <title>/description, exactly one rel=canonical, the og: block, and one application/ld+json graph near </body>.
2. Open https://yourstore.com/llms.txt — your business summary in plain text.
3. Run the page through Google's Rich Results Test — Product, breadcrumbs, organization and FAQ all validate.

7. Troubleshooting

8. FAQ

Hyvä? Yes — head tags are rendered server-side through Magento's page-config APIs, identical on Hyvä and Luma. No compatibility add-on to buy.

Will it slow my store down? No. The module emits zero frontend JavaScript — only head tags and one JSON-LD block. Lighthouse doesn't move.

Does it conflict with Magento's own canonical settings? No — if core canonicals are on, the suite replaces them at render time so exactly one canonical is ever emitted.

What is llms.txt? A plain-text business summary at your site root that AI crawlers (ChatGPT, Claude, Perplexity, Gemini) read to learn about your business — like robots.txt, but for answers instead of crawling.

Multi-store? Fully scoped — every setting per store view, hreflang across your store views, llms.txt per store.