Technical SEO audit tool that finds the problems and queues the fixes

XeoRank's technical SEO audit tool crawls your site, checks each page for confirmed technical and on-page problems, and puts every issue it confirms into a fix queue. You can read the results in the dashboard, or ask your AI assistant to run the audit and fix the issues; it asks you before each tool call, and WordPress writes stay off for a site until you turn them on.

The audit runs on the crawl described on the site crawler page. This page is about what the rules check, what happens to the issues afterwards, and where the audit stops.

Who it is for

• Founders and marketers who want to know what is wrong with their site and which problems to fix first.
• Agencies that audit client sites and then have to make the fixes, often on WordPress and Elementor.
• Developers who want audit results in a terminal or an AI coding assistant instead of another dashboard.

What the audit checks

The rules that raise issues sit in three groups: technical SEO, on-page and content quality. The crawl also reports broken links, redirect chains and errors in structured data a page already has. Each issue carries a severity, so a site-wide canonical problem does not sit at the same level as a long title.

Examples of what gets flagged on individual pages:

Some problems only show up across the whole site, so the audit also checks for:

• duplicate titles, duplicate H1s and exact duplicate content;
• canonical collapse, where most crawled pages declare the same canonical URL, which usually means a template is pointing every page at the homepage;
• sitemap URLs whose page declares a different canonical;
• titles over the display limit across the site.

Rules that know what kind of page they are reading

A rule that is right for a blog post is often wrong for a pricing page. A privacy policy does not need an author byline, and a 404 page should not be told its content is thin. XeoRank's rules are page-type aware, and it files only issues it can confirm from the HTML. Editorial checks such as author bylines, cited sources and thin content are not filed as issues on any page, so legal, pricing and error pages are not held to them. You get a shorter list with fewer findings to dismiss by hand.

In crawls started from the dashboard, pages that build their content with JavaScript are rendered in a headless browser when they need it, so the rules read the text and structured data that appear after scripts run instead of an empty app shell. Crawls your assistant starts through MCP read the raw HTML.

From issue to fix

Every issue goes into a fix queue for the site, and issues can be linked to a goal, such as more organic traffic to the services pages, so the work stays tied to the result you want.

There are three ways an issue gets fixed:

1. You fix it yourself, using the issue description and the list of affected URLs.
2. Your AI assistant fixes it. Connect the XeoRank MCP server to Claude, ChatGPT, Codex, Cursor or VS Code and ask it to work through the queue. On WordPress, with the free XeoRank Connector plugin, it can rewrite titles and meta descriptions, set canonical URLs and index settings in Yoast, Rank Math or All in One SEO, add schema, upload images and edit Elementor pages as native elements. It can also work on Framer projects and on code sites deployed through GitHub and Vercel.
3. A developer fixes it, for problems that live in server configuration or deep in a theme.

On WordPress, writes are off for each site until you turn them on. New pages and page clones are created as drafts, edits to an existing page are written to that page, each change to a page's content keeps a snapshot you can restore, and the assistant can send you a preview link for a draft before it is published.

A typical request looks like this:

"Crawl example.com and list the critical issues. Fix the missing meta descriptions on the service pages and add FAQPage schema to /pricing for the questions already on it. Show me each change before you publish."

Ways to run it

• Dashboard: add the site, start a crawl, and read the issues and the pages they affect.
• MCP: your assistant calls start_crawl to begin and get_crawl_results to read the issues.
• CLI and API: the xyle command-line tool (npm package @xyleapp/cli, named after the old brand) has crawl and site-crawl commands, and the REST API lives at https://api.xeorank.com. See the docs.

You or your AI agent start each crawl. Run one after a release, a migration or a batch of fixes to confirm the issues are gone.

To check a single page without signing up, the free tools include meta tag, canonical, H1, hreflang, sitemap and redirect checkers.

What it does not do

• It does not change server configuration. Security headers, HTTPS certificates, host or CDN redirect rules and server response time are for your developer or hosting provider.
• It does not audit pages behind a login, and it obeys robots.txt, so pages your robots.txt blocks are not checked.
• It does not analyse server log files or show how Googlebot crawls your site. For Google's own view of indexing, connect Search Console.
• It does not apply fixes by itself. Every change is a tool call your AI assistant makes, and your AI client asks you to allow it first.
• It does not promise rankings. Fixing technical problems removes obstacles; it does not decide where a page ranks.

Pricing

Starter is $19 a month for one site, with a 14-day free trial and no card needed. Pro is $49 a month and Growth is $109 a month. See pricing for what each plan includes.