README

Rankbeam is a Laravel SEO package built for production applications. The free MIT core resolves metadata through a layered precedence chain and renders canonical URLs, Open Graph and Twitter cards, a linked @id JSON-LD graph, XML sitemaps and crawler controls straight from your Laravel models and config.

composer require rankbeam/laravel-seo

Production auditing and monitoring (Pro) and the Filament admin UI ship as separate packages. This package never pulls them in.

Upgrading from fibonoir/laravel-seo v1? See UPGRADING.md — v2 renames the vendor and carves the old "full suite" down to this core; the analyzer, scanner, redirect manager, 404 monitor, and admin UI live on as separate packages (laravel-seo-filament, free; laravel-seo-pro, commercial).

What this package does

Area	Details
Meta resolution	`SEOResolver` merges six layers — config → global DB defaults → model-type defaults → route defaults → computed model values → explicit `seo_meta` values. Null never overwrites a lower layer.
Computed fallbacks	Title/description/image/robots derived from model attributes. Description candidates are configurable (`seo.computed.description_fields`), normalized (HTML stripped, entities decoded), and truncated at a word boundary (default 160 chars, no ellipsis). Per-model robots/`noindex` is built in — a `getSEORobots()` hook or an `is_indexable` attribute, overridable per page via `saveSEO(['robots' => …])`.
Rendering	`TagRenderer` outputs HTML (`@seo` Blade directives), structured arrays (Vue/React), or Inertia Head format. JSON-LD is emitted with `JSON_HEX_*` escaping so `</script>` in content cannot break out of the script element.
Canonical policy	Derived canonicals (model URL / current URL) get the query string stripped; explicitly set canonicals are preserved verbatim.
Schema (JSON-LD)	Builders for Article, Breadcrumb, FAQ, LocalBusiness, Organization, Product; `SchemaGraph` for Organization/WebSite/WebPage nodes cross-linked via stable `@id`s; breadcrumbs from a page's ancestor chain with a loop guard.
Sitemaps	`SitemapBuilder` (wraps spatie/laravel-sitemap) with config-driven model sources, programmatic named sources via `SEO::sitemaps()->register(...)`, sitemap index support, `seo:sitemap` command, and `/sitemap.xml` routes that can be disabled. Sitemaps open as a readable, branded page in the browser via a styled XSL stylesheet (on by default; search engines ignore it).
llms.txt	`seo:llms-txt` writes an optional markdown `llms.txt` index from the same sources as the sitemap (the registry + `seo.sitemap.models`), so the two never disagree. It's a compatibility file for tools that choose to consume it — Google Search does not use it. Served at `/llms.txt`, gated by config.
AI crawler control	`seo:robots-txt` renders a managed `robots.txt` (and `ai.txt`) from a doc-verified catalog of AI crawlers tagged by purpose — allow the AI-search and assistant crawlers (`ai_search`/`ai_assistant`), disallow the ones that train on you (`ai_training`) by default. `SEO::robotsTxt()->aiDirectives()` for a paste-able block; `SEO::aiCrawlers()` for the catalog + policy. Bots that ignore robots.txt are flagged advisory.
Markdown for bots	Content negotiation that serves clean markdown to AI crawlers instead of HTML — on `Accept: text/markdown`, `?format=md`, or (opt-in) a known AI crawler — from a model's `toSeoMarkdown()`, a `SEO::markdown()->register()` source, or a built title+description+content fallback. Off by default; never touches a normal visitor's response.
Indexing guard	Ties indexability to the Laravel environment: outside `seo.indexing_guard.allowed_environments` (default `['production']`) every page is forced to `noindex,nofollow` (above the whole precedence chain — overrides even a stored per-page value), `robots.txt`/`ai.txt` go disallow-all, and `seo:audit` warns. Stops a staging/local copy leaking into the index. Off by default, one-line opt-in `SEO_INDEXING_GUARD=true`, inert on production. See Indexing guard.
Generated OG images	Optional 1200×630 Open Graph images rendered by a real headless browser (`spatie/browsershot`) — correct multi-line wrapping, CJK/accents and truncation. Three publishable templates (default / article / product, selectable per model) + bundled OFL font; `seo:og-images` pre-generates and caches each card (content-hashed); the resolver serves it as a computed `og:image` fallback (existence-gated — never renders on a web request). Off by default, browser is a suggested dependency, static pre-generation only. See Generated OG images.
Warnings	`SEOWarningEvaluator` for admin UIs: title / description over the script's budget (60/160 for Latin, ~30/80 for CJK — see Multilingual content), manual-vs-fallback indicators, social-image dimension checks (min 200x200, ideal 1200x630, local files only).
Free audit	`seo:audit` — an in-process "what's wrong with my SEO right now" command (no queue, license, or network). Runs the metadata-class checks (missing / over- / under-length title & description, OG image, robots conflicts, canonical format/cross-domain/shared/insecure, hreflang code/duplicate/self-reference, focus keyword) and prints a per-page pass/warn/fail table with an explicit capability matrix. `--strict` for CI, `--json` for tooling. No numerical score (that's Pro).
Multilingual content	The package understands the language of your content, not only yours: per-script title/description budgets (`LengthPolicy`, graphemes not bytes), grapheme-safe truncation that cuts CJK at a sentence mark, locale-aware casing (Turkish İ/ı, Greek final sigma, ß), hreflang normalisation + self/x-default policies + audit codes, `inLanguage` on the schema graph, Yandex/Baidu/Naver/Seznam/Sogou/360/Cốc Cốc/DuckDuckGo in the crawler catalog, site-verification tags for eight engines, per-script OG-image font stacks with a missing-font pre-flight, IDN-tolerant canonical validation. Seventeen UI languages via `--tag=seo-lang`. See Multilingual content.
Explain resolution	`seo:explain "App\Models\Post" 42` — the resolver precedence trace: per field, the winning layer + value, the losing layers it overrode, and post-processing notes (title suffix, canonical strip, og:url derivation, image absolutization, indexing guard). Plus a site-level ledger (site name, default locale, canonical host) naming each value's source. Human `route:list`-style output + `--json`. Read-only; can't drift from what renders. See Explain resolution.
Migration importer	`seo:import-from ralphjsmit` — bulk-import SEO data from a competing Laravel package's storage into `seo_meta`. Idempotent, `--dry-run`, `--model=` scoping, explicit field mapping, morph rows re-resolved to the live model. See Migrating from other packages.
WordPress importer	`seo:import-from wordpress-csv` (a CSV export) and `seo:import-from yoast` / `rank-math` (the live WordPress DB via `--connection=`). Maps Yoast/Rank Math keys explicitly (incl. OG/Twitter overrides), resolves `%%title%%`-style template tokens, matches posts to your models by slug, and emits a redirects CSV for Pro. See Migrating from WordPress.

Database tables: seo_meta (per-model explicit values, morph + locale) and seo_defaults (global/model-type/route defaults). Nothing else.

Requirements

PHP 8.2+
Laravel 11, 12, or 13 (CI runs the full matrix; Laravel 13 requires PHP 8.3+)
spatie/laravel-sitemap ^7.0 or ^8.0 (suggested, required for sitemap generation)

Installation

composer require rankbeam/laravel-seo
php artisan vendor:publish --tag=seo-config
php artisan migrate

Quick start

use Rankbeam\Seo\Traits\HasSEO;

class Post extends Model
{
    use HasSEO;

    public function getUrlForSEO(): string
    {
        return route('posts.show', $this);
    }
}

<head>
    @seo($post)
</head>

Explicit values from the admin side:

$post->saveSEO([
    'title' => 'Custom SEO Title',
    'description' => 'Custom meta description',
]);

Headless / Inertia:

use Rankbeam\Seo\Facades\SEO;

return Inertia::render('Post', [
    'seo' => SEO::forInertia($post),
]);

Sitemap sources:

// config/seo.php
'sitemap' => ['models' => [Post::class => ['priority' => 0.8]]],

// or programmatically (e.g. in a service provider)
SEO::sitemaps()->register('pages', fn () => ['/about', '/contact']);

Then generate the files (requires spatie/laravel-sitemap) — the package's /sitemap.xml route serves what this command writes:

composer require spatie/laravel-sitemap
php artisan seo:sitemap

Serving your own static /sitemap.xml? Disable the package routes:

// config/seo.php
'routes' => ['enabled' => false],

Audit your SEO

# Audit the models under seo.audit.models / seo.sitemap.models
php artisan seo:audit

# Or target specific models, CI-fail on any issue, or emit JSON
php artisan seo:audit --model="App\Models\Post" --strict
php artisan seo:audit --json

A free, in-process pass/warn/fail report over the metadata-class checks. The rendered-HTML and live-canonical checks, and the numerical score, are part of the Pro scan — the command prints that boundary every run.

Generate llms.txt

php artisan seo:llms-txt            # writes public/llms.txt
php artisan seo:llms-txt --print    # print to stdout (dry run)

A markdown index of your site for AI crawlers (llms.txt), built from the same sources as your sitemap — registered sources plus seo.sitemap.models, with the same noindex/unpublished exclusions — so the two never disagree. It is served at /llms.txt (disable that route via seo.llms_txt.route), and you can schedule it alongside the sitemap:

Schedule::command('seo:llms-txt')->daily();

AI crawler control (robots.txt / ai.txt)

php artisan seo:robots-txt          # writes public/robots.txt
php artisan seo:robots-txt --print  # print to stdout (dry run)
php artisan seo:robots-txt --ai-txt # also write public/ai.txt

A managed robots.txt for the AI era, rendered from a doc-verified catalog of AI crawlers tagged by purpose. The default policy allows the AI-search and assistant crawlers (ai_search / ai_assistant) and disallows the ones that train on your content (ai_training); override per purpose or per bot in seo.ai_crawlers. Grab just the managed block for an existing file with SEO::robotsTxt()->aiDirectives(), or serve /robots.txt dynamically (off by default — it won't shadow a static file). Bots documented not to honour robots.txt are flagged advisory. See AI crawler control.

Guides

Long-form articles on the Rankbeam blog, a Laravel app that runs this package on itself:

Test status

Recorded full suite for core 3.20.0: 1036 passed (4363 assertions), 0 failed under PHP 8.4 / Laravel 13, with 17 default skips (16 opt-in Chrome image checks and one ext-xsl check). Real OG/PDF rendering and runtime-consumer checks run separately; default skips are not evidence that those paths ran. Public CI covers the supported PHP 8.2–8.4 / Laravel 11–13 combinations; Laravel 13 requires PHP 8.3 or newer.

git clone https://github.com/rankbeam/laravel-seo.git
cd laravel-seo
composer install
vendor/bin/pest

What is not in this package

Queued site scans, content analysis, redirect manager, 404 monitor, and the SEO dashboard ship in laravel-seo-pro (commercial); the Filament admin form fields ship in laravel-seo-filament (free). The old seo:install stub-publishing flow is gone.

Contributing

Translations are welcome — see TRANSLATING.md (language files, glossary, the parity test).

Bug reports, fixes and features are welcome — see CONTRIBUTING.md for the package boundary (what belongs in core vs. the Filament/Pro packages), how to run the suite, the rendering contract, and the versioning/BC policy.

Security

Please report security vulnerabilities privately, not in a public issue — see SECURITY.md.

License

MIT — see LICENSE.md.

rankbeam / laravel-seo

Package info

Security

Statistics

README

What this package does

Requirements

Installation

Quick start

Audit your SEO

Generate llms.txt

AI crawler control (robots.txt / ai.txt)

Guides

Test status

What is not in this package

Contributing

Security

License