Jump to

Skill · SEO site health audit

SEO site health audit.

Triage crawl findings by impact on traffic, not by the crawler's severity label.

Take a long list of crawler findings and triage it into a prioritized backlog mapped to business impact. Crawlers label issues critical, warning, or notice, but two issues both labeled critical can have wildly different real impact, so the audit triages on three axes: page-level traffic impact, mechanism of impact, and effort.

It runs on Ahrefs Site Audit results and looks for the clusters where one template change resolves thousands of issues at once. The discipline is fixing what moves traffic, then proving it did.

Audience: SEO and engineering teams facing a long crawler issue list, scoping a technical sprint, or deciding the highest-impact fixes when developer time is limited.

View the skill on GitHubBrowse the full catalog

The framework

Triage on three axes, not the severity label.

A 'critical' issue on a zero-traffic tag archive matters less than a 'warning' on a top-10 landing page. These three axes combine into a triage matrix yielding P0 to P3 or Park.

  1. 01Page-level traffic impact: tier the affected URLs as Tier 1 (top 10% of pages by organic traffic), Tier 2 (ranks but low traffic), or Tier 3 (exists, does not rank). A Tier-1 fix is worth roughly 10x a Tier-3 one.
  2. 02Mechanism of impact: does the issue actually move rankings? High mechanism (indexability, crawlability, renderability, Core Web Vitals, structured-data errors, internal-link integrity, hreflang) over low or theoretical (title and meta length warnings, decorative alt text, schema warnings).
  3. 03Effort: small (a config change or single template edit), medium (a component change or redirect-map work), or large (an architecture change, replatform, or schema overhaul).

Where the impact is

Clusters that resolve thousands of issues at once.

Most issues come in clusters with one root cause. These patterns produce outsized impact when fixed, so prioritize them even if individual issues look small.

  1. 01

    Redirect chain cleanup

    One sitemap update plus internal-link updates can resolve thousands of redirect-chain issues at once.

  2. 02

    Accidental noindex on a template

    A single line of code can re-index hundreds or thousands of pages that a template was silently blocking.

  3. 03

    Sitemap freshness pipeline

    A broken sitemap regeneration job affects every issue that depends on Search Console seeing the right URLs.

  4. 04

    Canonical on faceted navigation

    A template-level canonical fix resolves duplicate-content issues across an entire ecommerce category tree.

  5. 05

    Robots.txt restored

    A reverted production robots.txt can recover a sitewide drop within days.

  6. 06

    Hreflang block fix

    One template change resolves hreflang issues across a whole multilingual site.

  7. 07

    Image pipeline optimization

    Fixing the build process resolves thousands of individual asset issues in one change.

Runs on this platform

Where the crawl findings come from.

The triage works from Ahrefs Site Audit crawl results via the Ahrefs MCP. Ahrefs sees what its crawler finds, while Search Console reflects what Google actually indexes, so the audit reads both rather than trusting either alone.

Reference files

The reference that goes alongside the SKILL.md.

  • references/issue-impact-table.md

    A mapping table of common crawler issues to mechanism impact and typical fix effort, plus the triage matrix in detailed form.

Browse all reference files on GitHub

Bridges to other skills

Where the findings come from and go.

This skill triages a crawl. These cover the technical strategy behind the fixes, the single page, the specific-drop investigation, and the suite it feeds.

  • The strategy

    seo-technical

    Owns the technical strategy and schema design behind the fixes. This skill prioritizes a crawl's findings; that one defines what good technical SEO looks like.

  • A specific drop

    seo-traffic-diagnosis

    Diagnosing a specific traffic drop is a focused investigation, not a backlog triage. This audit prioritizes the whole crawl; that one chases one symptom.

  • One page

    seo-onpage

    When a fix is really about a single page's content and tags, that is on-page work. This audit operates at the site and template level.

  • The full audit

    seo-audit-orchestration

    The orchestrator runs this triage as one dimension of a coordinated suite. Reach for it when the brief is a whole-site baseline, not a technical backlog alone.

Open source under MIT

Read the SKILL.md on GitHub.

The skill source lives in the rampstackco/claude-skills repository alongside dozens of other skills covering the full lifecycle of brand and product work. This page is a structured overview; the SKILL.md is the source. MIT licensed.

View SKILL.mdBrowse the full catalog

Frequently asked questions.

How is this different from running a crawler?
The crawler produces the list of issues; this skill triages it into a prioritized backlog. Crawlers label issues critical, warning, or notice, and those labels are useful but not sufficient, because two issues both labeled critical can have wildly different actual impact on the property. Use the crawler tool directly to run the crawl, and this skill to decide what to fix first when a long list of issues is creating decision paralysis.
How are findings prioritized?
On three axes combined into a triage matrix. Page-level traffic impact tiers the affected URLs (Tier 1 is the top 10 percent by organic traffic, Tier 3 barely ranks). Mechanism of impact asks whether the issue actually moves rankings. Effort sizes the fix as small, medium, or large. A Tier-1, high-mechanism, small-effort issue is P0 (do this week); a Tier-3 low-mechanism issue is Park. P0 and P1 earn the team's attention, P2 goes on the roadmap, and P3 batches into routine maintenance.
What counts as a high-impact issue?
Issues with a real mechanism on rankings or traffic: indexability problems (accidental noindex, robots.txt blocks, canonical errors), crawlability problems (crawl traps, infinite redirects, 5xx errors, slow server response), renderability problems (JS errors blocking critical content), Core Web Vitals in the poor band, structured-data errors on rich-result-eligible pages, internal-link integrity (404s on internally linked URLs), and hreflang errors on multilingual sites. Lower-impact items include title and meta length warnings, a missing H1 on a low-traffic page, decorative-image alt text, and schema warnings rather than errors.
Why fix issues in clusters instead of one at a time?
Because most issues come in clusters that share a single root cause, so the right unit of work is the cluster, not the individual issue. A single template change can re-index thousands of pages, one sitemap update plus internal-link fixes can resolve thousands of redirect-chain issues, and a template-level canonical fix can clear duplicate content across an entire category tree. Fixing one redirect chain when 800 share the same root cause is busy work; group fixes that share an effort and resolve them together.
Why does 'we fixed 1,200 issues' miss the point?
Because counting issues without showing traffic or ranking impact wastes engineering credibility, and a clean technical report with no traffic gain helps no one. Define the metric that will prove a fix worked before shipping it, prioritize by traffic and mechanism rather than by issue count, and re-crawl after major fixes to confirm resolution. Severity worship (treating every 'critical' label as truly critical) and fixing low-mechanism issues for show are the failures this triage exists to prevent.