use-seo
Skill68 starsChanged 2 days agoMITView on GitHub
The SEARCH surface of a public page — the part a visitor never sees and a search engine judges the site by. Load it whenever you create a public route, change an address, touch metadata, add a section to the sitemap, or the owner says "the page is not in Google", "the wrong text shows up in the preview", "the picture is missing when I paste the link", "add hreflang", "set up robots". Every rule here was bought with a defect that looked like nothing: the page opens, the build is green, the gate passes — and whole languages are missing from results, or the page exists for nobody because no sitemap names it. Read it BEFORE the first file: a canonical address chosen wrong is not a cosmetic mistake, it is the page asking not to be indexed.
2 other projects use a version of this.
Adopted by
Most recently taken up by
- Fractera/web4-node-starter-kitword for word, first seen Nov 15, 2026
- Fractera/Agentic-Engineering-Infrastructureword for word, first seen Nov 8, 2026
Other versions2 projects use a version of this.
Word for word · 2
- Fractera/Agentic-Engineering-Infrastructure★ 68100% same
- Fractera/web4-node-starter-kit★ 68100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository91 other files it gives its agents.
- app/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- adopt-suggested-skillSkill
- ai-elementsSkill
- ai-sdkSkill
- answer-elementSkill
- ask-elementSkill
- audit-broken-charactersSkill
- build-product-with-ownerSkill
- content-factorySkill
- core-build-returnSkill
- core-scheduleSkill
- create-elementSkill
- create-multilingual-content-entrySkill
- dedicated-serverSkill
- design-and-settings-of-elementsSkill
- element-permissionSkill
- element-powerSkill
- expand-site-languageSkill
- explain-this-projectSkill
- find-skillsSkill
- fractera-vs-anthropicSkill
- handle-needSkill
- helpful-linksSkill
- help-in-browserSkill
- humanizerSkill
- integratorSkill
- interface-designSkill
- know-the-architectSkill
- manage-app-settingsSkill
- money-and-resourcesSkill
- node-typeSkill
- own-agentsSkill
- persist-env-var-with-rebuildSkill
- place-page-in-menuSkill
- provider-rulesSkill
- release-templateSkill
- researchSkill
- review-pageSkill
- route-to-elementSkill
- self-reviewSkill
- shadcnSkill
- shape-the-needSkill
- skill-creatorSkill
- study-for-elementSkill
- supervise-elementSkill
- translate-pendingSkill
- ultracode-sessionSkill
- use-agentic-ragSkill
- use-ai-generationSkill
- use-aioSkill
- use-app-configSkill
- use-auth-providersSkill
- use-authSkill
- use-browserSkill
- use-channelsSkill
- use-chatSkill
- use-code-shapeSkill
- use-databaseSkill
- use-dataSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 124
Keep the page's own illustration and its card picture the same value.
- Distinctive: line 147
All of them were green before your change too.
- Very specific: line 156
Read /sitemap.xml and /robots.txt on a deployment that has an address, and find your page there.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
use-seo
Informational, not binding. Know a better way for the case in front of you — do it your way and say so. You are trusted with the creative work on this project.
One builder per signal, and the page declares itself. Everything below follows from those two.
| Signal | The only builder | Never |
canonical + hreflang + the markdown twin link | buildAlternates(lang, subPath) — lib/seo/alternates.ts | a canonical written by hand, or inherited from the layout |
| the whole metadata object (title, description, OG, X card, robots, verification, icons) | constructMetadata() — lib/construct-metadata.ts | a hand-assembled Metadata |
| any address of our own page — in metadata, in a sitemap, anywhere | urlFor(lang, subPath) | a template string glueing site, language and path |
| structured data | three printers, §4 | a fourth printer of the same type |
A page built through createContentPage / createContentPost gets all four for free — that is what
the factory is for. A page written by hand owns all four itself.
1. 🔒 A page that does not declare itself gets the layout's declaration
Metadata in Next is inherited. [lang]/layout.tsx calls constructMetadata() with no pathname,
and that is deliberate: while it defaulted to /, every page that forgot its own alternates
announced the site root as its canonical — /ru literally said "index him, not me".
Which is why check:seo refuses a layout that declares alternates or pathname at all, and refuses
a page with no generateMetadata. A missing canonical is harmless (the page is read as its own
original); a canonical pointing elsewhere hands the page away, and one pointing at another domain
hands away the site. Never write a host into page data — the origin comes from APP-CONFIG.
2. One language means no language in the address
SINGLE_LANG_MODE turns on by itself when NEXT_PUBLIC_SUPPORTED_LANGUAGES holds one entry, and then
proxy.ts strips the segment from every public address: /en/blog answers 301 and the page lives at
/blog. That is the whole reason for the urlFor rule — a hand-glued address is a second source of
truth that disagrees with the first only in that mode, so it ships looking correct and turns every
canonical and every sitemap row into a redirect.
With two or more languages buildAlternates also emits x-default plus one entry per language. A
translation that does not name its siblings is not read as a translation but as a copy, and a set of
copies where nobody claims the original is what a search engine calls a doorway. The cost is not a
lower position — it is entire languages absent from results while the site looks perfect to its owner.
3. The snippet is not the first paragraph
description is cut to 160 characters (MAX_DESCRIPTION_LENGTH, then truncate adds an ellipsis).
It travels into the result page and into the social card, and that is a different job from the opening
paragraph, which explains the product to a human and takes the room it needs. Feed one field to both
and you get either a snippet cut mid-word or a hero of one line — the home page carries that reasoning
in its own _data/index.ts; do not undo it.
The five language-sensitive values (name, description, title template, keywords, site name) resolve
through configValueForLang. Without lang a Spanish page introduces itself in English.
4. Structured data has exactly three printers
| Who prints | What | Where |
[lang]/layout.tsx | WebSite, Organization, LocalBusiness — each behind its own cfg.jsonLd.* switch | builders in lib/jsonld.ts |
| the page factory | Article / NewsArticle, author = Person (@id, sameAs), publisher = Organization, and FAQPage | inline in lib/content/create-content-page.tsx |
components/nav/breadcrumbs.server.tsx | BreadcrumbList | beside the trail a human sees |
🔒 Breadcrumb markup lives where the drawn path lives, and nowhere else. The factory used to print
a second BreadcrumbList; two declarations of one path drift apart at the first edit, and silently,
because nobody opens the second one in a browser.
🔒 A switch can hide markup you correctly wrote. breadcrumbs in PLATFORM-CONFIG gates the
trail AND its BreadcrumbList together; faq does the same for the questions and FAQPage. Both
default to on, but an owner who turned one off has a site where your structured data is simply
absent — and it looks exactly like a page you built wrong. Before hunting your own mistake, read the
switch: npm run read:app-config and the platform config beside it. A requested feature that is
switched off is a sentence to the owner, not a bug to fix.
🔒 A switch is honoured at the SOURCE, not at each consumer. featureOn("faq") blanks the array
once, so the visible section and the FAQPage node cannot disagree. Gating only the visible half
promises a machine something the page does not contain.
Author and brand are read, never typed — lib/author.ts, lib/brand.ts, both from APP-CONFIG.
sameAs is what consolidates a person into one entity; two spellings split that entity in half. An
empty author is a legitimate state: no byline, no Person node. Better no attribution than a false one.
5. 🔒 A section named by no sitemap does not exist
Everything above guards the QUALITY of a page's signals and says nothing about a page nobody can find — and that error costs more. It shipped twice: the blog answered 200 with two translated posts while no map knew it, then all five footer pages, for a reason worth memorising —
check:seocounts a section as a first-level folder ofapp/[lang]holding apage.tsx. Bracket groups ((publicLayer),(footerPages)) are transparent to it. Transparency for the URL is not transparency for the map. A page inside a group is guarded by nobody — name it yourself.
| Set | Map | Why |
| finite, authored — home, sections, posts, footer pages | app/sitemap.ts | posts arrive from _list.generated.ts, so a new post enters the map by existing; footer pages are a literal list, so a private page cannot walk in on its own |
| grows at runtime — products | app/products/sitemap.ts, chunked | the limit is 50 000 URLs per file and exceeding it makes an engine drop the file whole; addresses here are rows × languages |
| behind a role | never | it does not exist at build, and a crawler would be handed the login form |
robots.ts is config-driven (seo.indexing, seo.disallowPaths, plus the service paths) and must list
every map, product chunks included: the storefront lazy-loads its rows, so the map is the only door
an engine has to them.
🔒 A sitemap that throws takes the whole build with it — map generation runs inside the page-data
phase. The product map wraps its query in try/catch and ships empty on failure; a deployment already
died this way.
6. The picture in the card
og:image must be absolute — Telegram, Facebook and LinkedIn ignore a relative path — and the file
must exist. Nothing checks the second half: check:content verifies assets named in sections, never in
meta. Five footer pages shipped pointing at /og-default.png, a file absent from the repository, and
the preview card of every one of them was empty. Request the URL before you believe it.
Keep the page's own illustration and its card picture the same value; two different ones mean the human
and the machine are looking at different pages. One file, three roles — heroImage under the
title, og:image + twitter:image, and Article.image — declared once in _data/meta.ts.
Where that file lives: public/page-media/<slug>.jpg, 1200×630. Posts keep theirs in
public/blog-media/, and public/placeholders/ belongs to the settings slots — do not park a page
picture in either. No text inside the picture: a word baked into pixels is not translated, and the
site has more than one language.
When the owner has no picture at all, /og-default.png answers — it is a ROUTE
(app/og-default.png/route.ts), not a file: a card drawn at build from the project's own identity,
shapes seeded by the site name, the owner's logo composited when he uploaded one. Never our mark,
never a letter of text. That address exists because five pages pointed at a file nobody had created.
7. What the gates say, and what they cannot
| Command | Refuses |
check:seo | a page with no generateMetadata; no alternates; openGraph without url; an address glued by hand inside a sitemap; a layout claiming a canonical; a first-level section present in no map |
check:aio | a public page with no markdown twin — see use-aio |
check:content | the link and asset rules of the body — use-links owns them, and they are part of this surface: an outgoing link spends the weight this page earned |
npm run build | the honest answer to "is it still static": ●, never ƒ |
All of them were green before your change too. Green proves the tree is well-formed, never that the page is findable.
8. Before you call it done
- Fetch the built page and read its head: one
canonical, oneog:url,hreflangfor every enabled language plusx-default, thealternateof typetext/markdown, and theld+jsonsections — count them, a duplicatedBreadcrumbListis invisible any other way. - Request the
og:imageURL itself. Empty card, missing file, same thing. - Read
/sitemap.xmland/robots.txton a deployment that has an address, and find your page there. - 🔒 An empty sitemap on your own machine is NORMAL, not a defect.
app/sitemap.tsopens withif (!site) return [], andsitecomes fromAPP-CONFIG, which stays{}until a server fills it. Judge the map where there is an address, never here.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

