菜单
Technical audit knowledge for XML sitemaps: discovery, structural checks, URL quality validation, lastmod integrity, CMS-specific patterns, and sampling. Use when evaluating, diagnosing, or fixing a sitemap — including sitemap index structures, sub-sitemap issues, and robots.txt conflicts.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | sitemap |
| description | Technical audit knowledge for XML sitemaps: discovery, structural checks, URL quality validation, lastmod integrity, CMS-specific patterns, and sampling. Use when evaluating, diagnosing, or fixing a sitemap — including sitemap index structures, sub-sitemap issues, and robots.txt conflicts. |
Based on Google's sitemap specification (sitemaps.org protocol) and observed patterns across WordPress, PrestaShop, Shopify, and Magento sites.
Sitemap: directive — authoritative. Sitemap: https://domain.com/sitemap_index.xml/sitemap_index.xml (WordPress with Yoast / Rank Math)/sitemap.xml (generic)/1_index_sitemap.xml (PrestaShop default — the standard path does not exist)/sitemap/ (some custom implementations)Not declaring the sitemap in robots.txt is a low-severity issue, not critical.
Google can still find sitemaps via Google Search Console submission or standard paths.
Best practice: always declare it with Sitemap: https://domain.com/sitemap_index.xml.
| Type | Root element | Contains |
|---|---|---|
| Sitemap index | <sitemapindex> | References to sub-sitemaps (<sitemap><loc>) |
| URL set | <urlset> | Individual URLs (<url><loc>) |
| News sitemap | <urlset> + news namespace | <news:publication>, <news:publication_date> |
| Image sitemap | <urlset> + image namespace | <image:image>, <image:loc> |
Most CMS plugins generate a sitemap index pointing to type-specific sub-sitemaps (pages, posts, products, categories). This is the recommended structure for sites with more than a few hundred URLs.
/1_index_sitemap.xml. Without a 301 redirect from
/sitemap.xml, the standard path returns 404.Most impactful and frequently missed sitemap error. The sitemap file returns HTTP 200
but has X-Robots-Tag: noindex in its response headers. Google reads the header, treats
the sitemap as noindexable, and stops processing it — even though the status is 200.
How to detect: check response headers of the sitemap URL directly, not just the HTTP status.
Causes:
Fixes by environment:
LiteSpeed Cache:
Advanced > Excludes > Do Not Cache > add /sitemap*.xml
Apache (.htaccess):
<FilesMatch "sitemap.*\.xml$">
Header unset X-Robots-Tag
</FilesMatch>
Nginx:
location ~* sitemap.*\.xml$ {
more_clear_headers "X-Robots-Tag";
}
If it persists: server-level configuration — contact hosting support.
A parse error in the XML makes the sitemap unreadable. Causes:
& instead of &, < in URLs)Validation: https://www.xml-sitemaps.com/validate-xml-sitemap.html
Google's hard limit is 50 MB uncompressed. Practical warning threshold: 30 MB.
If approaching the limit: switch to a sitemap index structure and split by content type (pages, posts, products, categories). Each sub-sitemap can contain up to 50,000 URLs.
The sitemap should be served with Content-Type: application/xml or text/xml.
Serving it as text/html or application/octet-stream can cause parsing issues
in some crawlers.
A slow sitemap reduces how frequently Googlebot re-fetches it.
Fix: enable sitemap caching in the SEO plugin, or serve the XML file from static cache.
A sub-sitemap referenced in the index that returns 404 or error causes Google to lose all URLs from that sub-sitemap. The index references it but it does not exist.
Fix: regenerate the sitemap from the SEO plugin, or remove the broken reference.
Sub-sitemaps with 0 URLs generate unnecessary Googlebot requests and waste crawl budget.
Fix in WordPress: Yoast SEO > XML Sitemaps > disable the post type or taxonomy that generates the empty sub-sitemap.
Author archive pages are typically thin content (a list of posts by one author, with no unique content of its own). Having them in the sitemap signals them as indexable and wastes crawl budget.
Fix:
URLs in the sitemap starting with http:// on an HTTPS site are incorrect.
Google may not index them correctly as it treats them as different URLs from
the canonical HTTPS versions.
Fix: check WordPress > Settings > General > Site Address (URL) — must be https://.
Regenerate sitemap after correcting.
A sitemap mixing https://domain.com/ and https://www.domain.com/ URLs indicates
an incorrect canonical configuration or a sitemap generated before/after a migration.
Fix: define the canonical version in WordPress > Settings > General, or in the SEO plugin. Regenerate sitemap.
If more than 10% of URLs disagree on trailing slash (some end with /, others do not),
it signals incorrect canonical configuration. Google treats these as distinct URLs.
Fix: enforce consistent trailing slash in permalink settings. Yoast applies the WordPress permalink configuration automatically.
/Pagina/ and /pagina/ are treated as different URLs by Google. URLs in sitemaps
should always use lowercase paths.
Fix: normalize to lowercase. In WordPress, use a redirect plugin (Redirection) to force lowercase. Update slugs directly in the CMS.
URLs with utm_source, utm_medium, utm_campaign, fbclid, gclid, etc. in the
sitemap are indexed as independent URLs with the same content as the canonical version
— creating duplicate content.
Fix: configure the SEO plugin to exclude URLs with tracking parameters from the sitemap.
Staging domain signals: staging., dev., .local, test., preprod., qa., sandbox.
If staging URLs appear in the production sitemap, the sitemap was generated or copied from a staging environment. Regenerate the sitemap in production.
URLs pointing to a different domain than the audited site. Typical causes:
URLs exceeding 2,000 characters can cause issues in some crawlers. Shorten slugs or remove unnecessary URL parameters.
A URL present in the sitemap but blocked by a Disallow rule in robots.txt creates
a contradiction: the sitemap says "please index this URL" but robots.txt says
"do not crawl it." Google will see the URL but not be able to crawl it.
Fix: either remove the Disallow rule for URLs that should be indexed, or remove those URLs from the sitemap.
lastmod tells Googlebot when a page was last modified, helping it prioritize
re-crawls of recently updated content.
At least 50% of URLs should have a lastmod value. Sites without lastmod give
Googlebot no signal for prioritizing re-crawls.
Enable in WordPress:
Documented bug: Rank Math generates 1970-01-01T00:00:00+00:00 for posts
that do not have a saved modified date in the database (typically posts published
before Rank Math was installed, or posts that were never re-saved).
This is not a minor cosmetic issue — Google sees it as a last-modified date in 1970 and de-prioritizes those URLs for re-crawl.
Fix:
Valid formats (W3C datetime / ISO 8601):
YYYY-MM-DD — date only, acceptableYYYY-MM-DDThh:mm:ss+TZ — full datetime with timezone, preferredInvalid examples: DD/MM/YYYY, Month DD, YYYY, timestamps without year.
Google ignores future modification dates. Cause: incorrect server timezone configuration. Fix: verify server time and timezone, then regenerate sitemap.
If more than 90% of URLs have a lastmod older than 2 years, the sitemap is not reflecting the real state of the content. Either:
If all URLs share the exact same lastmod date, the sitemap is statically generated with a single timestamp — not reflecting actual page modification dates.
Fix: enable dynamic lastmod generation in the SEO plugin.
The priority field (0.0–1.0) is largely ignored by Google and has been
de-prioritized in the sitemaps.org specification. Google's own guidance states
it does not use priority for ranking.
Common misuse: all URLs set to priority=1.0. When everything has maximum
priority, the signal is noise.
Recommendation: if the SEO plugin sets it automatically, leave it. If it is
forcing 1.0 everywhere, either configure differentiated values or remove it entirely.
Suggested values (if used):
| Page type | priority |
|---|---|
| Homepage | 1.0 |
| Main service / product pages | 0.8 |
| Blog / listing pages | 0.6 |
| Archive / tag pages | 0.4 |
| Paginated pages | 0.2–0.4 |
Sampling 10 random URLs from the sitemap reveals live issues that static analysis misses.
Even a small % of broken URLs in the sitemap wastes crawl budget and signals to Google that the sitemap is not maintained.
For comprehensive broken URL detection: use Screaming Frog in list mode with all sitemap URLs. Sampling only catches a statistical signal.
The sitemap should contain the canonical final URL — not a URL that redirects to it. If more than 20% of sampled URLs return a 3xx, the sitemap was not regenerated after a permalink change or migration.
Note: a single redirect is acceptable and Google follows it. The issue is having the redirected URL in the sitemap rather than the destination.
Pages with <meta name="robots" content="noindex"> or X-Robots-Tag: noindex
should not be in the sitemap. Including them creates a contradiction between
the sitemap signal (index this) and the page directive (do not index this).
Google's behavior: Google respects noindex over the sitemap. The URL will
not be indexed, but it will still be crawled — wasting crawl budget.
Common causes:
The URL in the sitemap should match the canonical URL declared on the page. A mismatch means the sitemap is pointing to a non-canonical URL.
Example:
https://domain.com/product-namehttps://domain.com/products/product-name/Fix: align permalink structure, canonical configuration in the SEO plugin, and regenerate the sitemap.
Pages consistently taking more than 2 seconds to respond will be crawled less frequently. Googlebot has a crawl budget and allocates less time to slow sites.
Diagnose: check TTFB in PageSpeed Insights. Activate page caching and CDN.
These URL patterns should never appear in a sitemap:
| Pattern | Why |
|---|---|
/wp-admin/, /wp-login.php | Admin — should never be crawled |
/feed/, /*/feed/ | RSS feeds — no SEO value in sitemap |
/cart/, /checkout/, /my-account/ | Transactional — personalized, no indexable value |
/?s=, /search/ | Internal search results — duplicate content, no value |
PHPSESSID=, id_lang=, id_currency= | Session / language parameters |
/wc-api/, route=checkout, route=account | WooCommerce / OpenCart internal endpoints |
utm_source=, fbclid=, gclid= | Tracking parameters |
?preview=true, ?page_id= | WordPress preview URLs |
/trackback/ | WordPress trackback endpoints |
/sitemap_index.xml/sitemap_index.xmllastmod=1970-01-01 for posts without a saved modified dateWooCommerce adds post types (product, product_variation) and taxonomies
(product_cat, product_tag) that require specific sitemap decisions.
/my-account/ in the sitemap
One of the most common WooCommerce sitemap errors. The My Account page is personalized per user and has no indexable value — yet SEO plugins often include it because it is a standard WordPress page.
Product variations
WooCommerce creates a post for each variation (product_variation post type).
These are not publicly accessible URLs — they return 404 if visited directly.
Some sitemap configurations accidentally include them.
Fix: Yoast > XML Sitemaps > verify product_variation is disabled.
Rank Math > Sitemap > Post Types > disable Product Variation.
Out-of-stock and discontinued products
Products set to "Draft" or with stock status "Out of stock" may remain in the sitemap after the product is discontinued. If the product page returns 404 or redirects, it should be removed from the sitemap.
Pattern to detect: sampling returns 404 or redirect on product URLs. Fix: in WooCommerce, set discontinued products to "Draft" or delete them and add a proper 301 redirect to a relevant category page.
Product tag archives
product_tag archives are typically thin content (a filtered list of products by tag,
with no editorial content). Equivalent to the author archive problem.
Fix: Yoast > Search Appearance > Taxonomies > Product Tags > No index. Rank Math > Titles & Meta > Product Tags > No index. Exclude from sitemap after setting to noindex.
WooCommerce endpoint pages
WooCommerce generates virtual endpoint URLs appended to the My Account page:
/my-account/orders/, /my-account/downloads/, /my-account/edit-account/, etc.
These should never be in the sitemap — they require user authentication.
These endpoints are not standard WordPress pages so they are less likely to appear in the sitemap, but verify if the site uses a custom page structure.
/shop/ page
The main shop page (/shop/ or localized equivalent) is generally worth indexing.
However, paginated versions (/shop/page/2/, /shop/page/3/) are low-value and
should be excluded from the sitemap (set to noindex via SEO plugin).
WooCommerce REST API and AJAX endpoints
/wc-api/, /wp-json/wc/ endpoints should never appear in a sitemap.
If detected, the sitemap is likely including URLs from a crawl-based generator
rather than from the WooCommerce post type registry. Switch to the native
Yoast or Rank Math sitemap generator.
/1_index_sitemap.xml — the path /sitemap.xml does not exist.htaccess from /sitemap.xml to /1_index_sitemap.xml?controller=...)
These must be excluded from the sitemap module config or blocked in robots.txt&p=2) — configure the module to exclude them/sitemap.xml (platform-generated, not editable directly)<meta name="robots" content="noindex">
on the page, which causes Shopify to remove it from the sitemap on the next regenerationcurl -s -I https://example.com/sitemap.xml | head -5
Resultado esperado: HTTP/2 200 y Content-Type: application/xml o text/xml.
Si devuelve 301 → hay redirect innecesario (Yoast genera esto a veces hacia el índice).
Si devuelve 404 → el sitemap no existe en esa ruta — verificar robots.txt para la ruta real.
# Total de <loc> en un sitemap o sitemap index
curl -s https://example.com/sitemap.xml | grep -c "<loc>https"
# Por sub-sitemap
curl -s https://example.com/post-sitemap.xml | grep -c "<loc>https"
curl -s https://example.com/page-sitemap.xml | grep -c "<loc>https"
# Buscar si una slug o URL concreta aparece en el sitemap
curl -s https://example.com/page-sitemap.xml | grep -i "nombre-del-slug"
Útil para confirmar si una página noindex fue eliminada correctamente del sitemap, o para verificar que una URL nueva ya aparece después de un ping/regeneración.
curl -s https://example.com/robots.txt | grep -i "sitemap"
CRITICAL (Google cannot process the sitemap)
[ ] Sitemap exists and is accessible (HTTP 200)
[ ] No X-Robots-Tag: noindex on the sitemap response headers
[ ] XML is valid and parseable
[ ] No staging / development URLs in production sitemap
[ ] No URLs pointing to a different domain
HIGH (crawl budget waste or indexation loss)
[ ] No sub-sitemaps returning 404 or error
[ ] URL count within the 50,000 limit per file
[ ] No URLs blocked by robots.txt Disallow rules
[ ] No HTTP URLs on an HTTPS site
[ ] No www / non-www inconsistency
[ ] No UTM or tracking parameters in sitemap URLs
[ ] No internal / transactional URLs (cart, checkout, admin, search)
[ ] WooCommerce: /my-account/ not in sitemap
[ ] WooCommerce: product_variation post type excluded from sitemap
[ ] WooCommerce: product_tag archives noindex and excluded from sitemap
[ ] No lastmod 1970-01-01 (Rank Math bug)
[ ] Sampling: no broken URLs (404, error)
[ ] Sampling: no noindex URLs in sitemap
[ ] Sampling: no canonical mismatch
MEDIUM (reduced crawl efficiency)
[ ] File size under 30 MB
[ ] Response time under 3 seconds
[ ] No trailing slash inconsistency
[ ] No uppercase in URL paths
[ ] No empty sub-sitemaps
[ ] Author archives not in sitemap
[ ] lastmod format valid (W3C / ISO 8601)
[ ] No future lastmod dates
[ ] lastmod not all identical (not statically generated)
[ ] lastmod not all older than 2 years (>90% threshold)
[ ] Content-Type is application/xml or text/xml
[ ] Sampling: redirect rate under 20%
[ ] Sampling: average response time under 2 seconds
LOW (maintenance and best practices)
[ ] Sitemap declared in robots.txt
[ ] No duplicate URLs in sitemap
[ ] No URLs longer than 2,000 characters
[ ] priority field not set to 1.0 for all URLs
[ ] No empty sub-sitemaps
Genera un informe de auditoría SEO en formato .docx para cotizaciones y propuestas comerciales. Recopila datos mínimos del sitio (volumen, stack, issues críticos), calcula horas de trabajo y genera un presupuesto en USD. Output: documento entregable al cliente. Invocar cuando el usuario pida "cotización SEO", "presupuesto SEO", "propuesta SEO", "informe de auditoría para cliente" o "/seo-quote".
Audita un sitio web frente a la California Consumer Privacy Act (CCPA, 2018) modificada por la California Privacy Rights Act (CPRA, 2020, en vigor desde 1 enero 2023). Detecta si el sitio cumple los umbrales de aplicación, evalúa los 8 derechos del consumidor, los requisitos de opt-out y los acuerdos con proveedores de servicios. Produce informe de cumplimiento con score e issues priorizados con cuantificación de sanciones en USD. Sub-skill del orquestador /privacidad — también invocable directamente.
Audita un sitio web frente a la Ley 25.326 de Protección de Datos Personales (Argentina). Detecta el sector automáticamente. Produce informe de cumplimiento con score, tabla de estado por artículo e issues priorizados con cuantificación de sanciones. Sub-skill del orquestador /privacidad — también invocable directamente.
Audita un sitio web frente a los requisitos de la Ley 21.719 de Protección de Datos Personales (Chile). Detecta el sector del sitio automáticamente y aplica requisitos específicos. Produce informe de cumplimiento con score, tabla de estado por artículo e issues priorizados.
Audita un sitio web frente a la Ley Federal de Protección de Datos Personales en Posesión de los Particulares (LFPDPPP — México, 2010) y su Reglamento. Detecta el sector automáticamente. Produce informe de cumplimiento con score, tabla de estado por artículo e issues priorizados con cuantificación de sanciones. Sub-skill del orquestador /privacidad — también invocable directamente.
Audita un sitio web frente a la Lei Geral de Proteção de Dados Pessoais (LGPD — Lei 13.709/2018, Brasil). Detecta el sector automáticamente. Produce informe de cumplimiento con score, tabla de estado por artículo e issues priorizados con cuantificación de sanciones en BRL. Sub-skill del orquestador /privacidad — también invocable directamente.