SEO

Every page gets a <title>, meta description, canonical link, Open Graph and Twitter card tags, hreflang links and optional JSON-LD, printed once by {{ seo_tags() }} in the layout.

Setting page metadata

From a controller:

use Spatie\SchemaOrg\Schema;

$seo = $this->app->seo
    ->title('About')
    ->description('Who we are.')
    ->image('/images/about.jpg');
$seo->schema(Schema::organization()->name('Acme')->url($seo->url('/')));

Or from a template, in the seo block (it renders before the tags are printed):

{% block seo %}{% do seo.title('About').description('Who we are.') %}{% endblock %}

Setter	Effect
title(?string)	<title> becomes "Title · Site name"; null means just the site name
description(?string)	meta description, og:description, twitter:description
canonical(string)	canonical URL (defaults to the current path, without the query string)
image(?string)	share image: a /path or an https:// URL; switches to the large Twitter card
type(string)	og:type: website (default) or article
article(published, modified, tags)	og:type=article plus article:* tags
noindex(bool)	keeps the page out of search engines
schema(Type)	adds a JSON-LD block built with spatie/schema-org

Site-wide defaults (name, description, default image, author) come from site in config/app.php.

Automatic metadata

• Blog posts are mapped from front matter by TinyTank\Blog\PostSeo: title, summary, image, og:type=article with dates and tags, and a BlogPosting JSON-LD block. Drafts get noindex.
• Error pages get noindex.
• hreflang links list the page in every language (or only where it exists, see Languages), plus x-default for the default language.

Sitemap, feeds and robots.txt

Three framework controllers, wired in config/routes.php:

URL	Controller	Content
/sitemap.xml	TinyTank\Seo\SitemapController	every static GET page in every language, plus every published post version
/blog/feed.xml	TinyTank\Blog\FeedController	Atom feed of the 20 latest posts in the current language (/it/blog/feed.xml…)
/robots.txt	TinyTank\Seo\RobotsController	allows everything and points to the sitemap

"Static" pages are routes without placeholders or a file extension. Drafts and untranslated posts never appear.

Absolute URLs come from APP_URL

Canonical URLs, og:url, og:image, the sitemap and the feeds use APP_URL, never the request's Host header. Pages are publicly cacheable, so a forged Host must not end up in a cached page. Set APP_URL correctly on every server.

JSON-LD is encoded so that a </script> in any value can't break out of the tag.