---
name: seo-audit
description: Audit supplied pages, crawl exports, or a reachable site for technical and on-page SEO issues, prioritizing observed problems and separating unavailable checks.
---

# SEO audit

Establish the site, business goal, scope, and available evidence. Use supplied exports first; fetch pages when needed and authorized. Record the response status and content type. A challenge, error, login screen, or HTTP 202 response is not a successfully inspected page.

Work from crawl access and indexability to page content and navigation. Keep these distinctions explicit:
- Robots.txt controls crawling; a disallow is not a reliable removal-from-index directive.
- A noindex directive requires Google to access it; blocking crawling can prevent that.
- A canonical is a signal, not a guaranteed instruction. Check its target, page intent, status, and conflicting signals.
- A sitemap lists candidate URLs; it does not prove indexation or incoming links.
- A successful response alone does not prove rendered content or Google indexation.

For pages in scope, inspect titles, headings, useful content, canonical and robots signals, internal links, relevant structured data, and usability. Prioritize defects by user/business impact, reach, confidence, and effort. Preserve intentional staging, account, and utility-page exclusions. Do not invent missing tags when HTML is unavailable.

Keep field performance measurements separate from lab measurements. Distinguish a Search Console observation from a hypothesis based on page HTML. A patent, third-party score, or SEO heuristic is not proof of a current ranking mechanism. Verify changing Google requirements against primary documentation.

Return a short prioritized table: issue, affected URL, observed evidence, recommended change, and verification step. List unavailable checks and the specific input needed. Avoid an arbitrary overall score, exhaustive generic checklist, or promised ranking gain.

[Google crawling and indexing guidance](https://developers.google.com/search/docs/crawling-indexing)