Website Care · Starter

Check that search engines can find the right pages

Traffic dropped, a redesign or move is coming, or you are not sure search engines see the pages you care about.

  • Any
  • Claude Code
  • Grok Build
  • Cursor
  • Codex CLI

For: Owner, Office manager, Tech person, Anyone at work · Updated November 10, 2026

A starting point: fill in the [brackets] with your own details, and check the answer before you use it. Tool screens change often, so yours may look a little different.

The prompts

Copy a quick prompt, or grab the full skill below.

QUICK PROMPT 1
Why this prompt works: Ask for a report first: indexing mistakes are costly to undo.
Check whether search engines can find and index the main pages of the site I work on. Flag anything blocked or marked 'noindex' by mistake. Plain-language report first, no changes.
QUICK PROMPT 2
Why this prompt works: Defining what counts as a problem (missing, duplicated, cut off) gives a clean checklist.
Review page titles and descriptions on the site I work on. List any that are missing, duplicated, or cut off, with the page each is on. Don't rewrite them yet.
QUICK PROMPT 3
Why this prompt works: Scope: tying the check to the move keeps it on what matters right now.
We're moving the site I work on to new URLs. Check that each old address points to the right new page, and list any that don't, before making redirect changes.

What this covers: Titles, descriptions, canonicals, robots and noindex safety, sitemaps, structured data, status codes, and checks before and after a site move.

For anyone improving a site or app, whether a small-business owner or someone at a larger company practicing better prompts. If search engines can’t find or understand the right pages, good content goes unseen, and one wrong setting can hide a whole site. Paste a prompt into any AI assistant that can see the site or repo.

Two ways to use it

  1. Any AI chat (Copilot, ChatGPT, Claude, Grok): copy one of the quick prompts above. Each one asks for a plain-language report first, so nothing changes until you say so.
  2. A coding agent working in the site’s code (Claude Code, Grok Build, Cursor, Codex CLI): save the full skill below as SKILL.md in a folder named seo-technical-audit, for example .claude/skills/seo-technical-audit/SKILL.md (Grok Build also reads that location, or use .grok/skills/). Then ask the agent to use the seo-technical-audit skill. It audits first, makes the smallest safe changes in a review branch, and ends with a report.

Download SKILL.md · All Website Care skills

The full skill (SKILL.md)

Last reviewed Nov 10, 2026. Standards and tools change, so check anything dated before relying on it.

---
name: seo-technical-audit
description: Use when asked to check or fix a website's technical SEO - titles and meta descriptions, canonicals, robots.txt and meta robots, XML sitemaps, structured data, redirects, 404/410 handling, internal linking, and making sure staging/preview copies are noindexed - especially before or after a migration or redesign. Read-only audit first, fixes as a review PR. Not for writing new content or keyword strategy.
---

# Technical SEO audit

Last reviewed: 2026-11-10

## Quick prompts

For anyone improving a site or app, whether a small-business owner or someone at a larger company practicing better prompts. If search engines can't find or understand the right pages, good content goes unseen, and one wrong setting can hide a whole site. Paste a prompt into any AI assistant that can see the site or repo.

1. "Check whether search engines can find and index the main pages of the site I work on. Flag anything blocked or marked 'noindex' by mistake. Plain-language report first, no changes."
   *Why this prompt works:* Ask for a report first: indexing mistakes are costly to undo.

2. "Review page titles and descriptions on the site I work on. List any that are missing, duplicated, or cut off, with the page each is on. Don't rewrite them yet."
   *Why this prompt works:* Defining what counts as a problem (missing, duplicated, cut off) gives a clean checklist.

3. "We're moving the site I work on to new URLs. Check that each old address points to the right new page, and list any that don't, before making redirect changes."
   *Why this prompt works:* Scope: tying the check to the move keeps it on what matters right now.

Make sure search engines can crawl, understand, and index the right pages, and nothing else. Technical only: no rewriting of copy.

## Hard limits

- Do not change page content, titles that the owner wrote, or URL structure without approval. Propose title/description fixes as a list for the owner unless they are clearly broken (empty, duplicated template text).
- Never remove `noindex` from a staging/preview build, and never add `noindex` or `Disallow: /` to production, without explicit approval. Getting this wrong de-indexes a site.
- Search Console, DNS verification, and sitemap submission are the owner's accounts; propose, do not do.
- Crawl politely (browser UA, pauses); review PR only.

## Inputs to gather

1. Production domain, staging/preview domains, and how each sets indexing (env flag, header, meta tag, robots.txt).
2. URL inventory: sitemap(s), CMS export, or a crawl. For migrations, the old site's URL list.
3. Search Console access (owner) for coverage and page indexing reports, if available.
4. Content types and templates.

## Step 1: Classify

| Type | Priorities |
|---|---|
| Article / blog | `Article`/`BlogPosting` structured data, author and dates, canonical on paginated archives, tag/category thin pages, image alt and dimensions, RSS |
| Marketing / landing | Unique titles per page, `Organization` data, no orphaned campaign pages, UTM URLs canonicalized |
| Docs | Versioned docs canonical to latest, `BreadcrumbList`, search-result pages `noindex` |
| E-commerce | `Product` + `Offer` data, faceted URLs (canonical or `noindex`), out-of-stock handling, pagination |
| App / SaaS behind login | Only marketing and public pages indexable; app routes `noindex` or blocked; no session IDs in URLs |
| SPA | Content and links present in rendered HTML (SSR/prerender preferred), real `<a href>` links, correct status codes for missing routes (not soft 404s) |

## Step 2: Check

- **Status codes**: every sitemap URL returns 200; removed pages return 404 or 410 (or 301 to a true equivalent); no soft 404s (200 pages saying "not found").
- **Titles and descriptions**: present, unique, roughly 50–60 and 120–160 characters as guidance, not truncated template text.
- **Headings**: one `h1` that matches the page topic; logical order.
- **Canonicals**: absolute URL, self-referencing on canonical pages, consistent with redirects (https, www or not, trailing slash), never pointing to a redirect or 404.
- **Robots**: production `robots.txt` allows important paths and lists the sitemap; staging/preview sends `X-Robots-Tag: noindex` or meta `noindex` (robots.txt alone does not de-index). Note: a page blocked in robots.txt cannot show its `noindex`.
- **Sitemaps**: only canonical, indexable, 200 URLs; accurate `lastmod`; under 50,000 URLs/50 MB per file.
- **Structured data**: JSON-LD matches visible content; validate with the Rich Results Test / Schema Markup Validator.
- **Internal links**: no links to redirects or 404s, no orphan pages, descriptive anchor text.
- **International** (if any): `hreflang` pairs reciprocal and canonical.
- **Media**: image `alt`, dimensions, modern formats; video pages have thumbnails and structured data if they matter.
- **Performance and mobile**: link to performance-web-vitals-pass; check viewport meta.
- **Migrations**: every old URL maps to a new one with a single 301 (see link-and-redirect-integrity-check).

## Step 3: Fix

Template-level fixes (meta tags, canonical logic, sitemap generation, structured data component, environment-based `noindex`) in one review PR. Content-level suggestions as a list for the owner.

## Validation

- Rebuild and re-crawl the preview: status codes, canonicals, robots, sitemap contents.
- Confirm preview stays `noindex` and production config (checked from source, not by deploying) would be indexable.
- Structured data validates without errors on one page per template.

## Report format

1. Interface type(s), URL counts (sitemap vs crawl vs inventory) and differences.
2. Findings table: issue, affected URLs/templates, evidence, fix, needs-approval.
3. Indexing safety check: staging and production states.
4. Owner to-do (Search Console, sitemap submission, content edits). PR link, "not merged".

## References

- Google Search Central SEO Starter Guide: https://developers.google.com/search/docs/fundamentals/seo-starter-guide
- Canonicalization: https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls
- Robots meta and X-Robots-Tag: https://developers.google.com/search/docs/crawling-indexing/robots-meta-tag
- robots.txt intro: https://developers.google.com/search/docs/crawling-indexing/robots/intro
- Sitemaps: https://developers.google.com/search/docs/crawling-indexing/sitemaps/build-sitemap
- Structured data intro and Rich Results Test: https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data , https://search.google.com/test/rich-results
- Site moves with URL changes: https://developers.google.com/search/docs/crawling-indexing/site-move-with-url-changes
- JavaScript SEO basics: https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics

Check before you send

  • It gave a plain-language report before suggesting or making any change
  • Each finding names the exact page or file and why it matters
  • It said what it could not check or verify
  • Nothing was changed, published, or deleted without your OK
  • Fixes are small and keep the existing design and content

Privacy note: what not to paste

Paste page addresses, screenshots, or reports, not passwords, API keys, customer data, or private admin links. Only check sites you or your employer control, and get the site owner’s OK before anyone changes anything.