Log inSubmit server

Technical SEO Site Audit

Firecrawl + Chrome DevTools + Search Console

Crawl a site’s key templates, check titles, canonicals, structured data, internal links and rendering, compare them with Search Console indexing data, and get a prioritized fix list tied to real traffic.

3 MCP servers combined

Included MCP Servers (3)

  • Firecrawl MCP (official)

    Maps the site’s URLs and scrapes pages to markdown and HTML for on-page checks.

    npx -y firecrawl-mcp
  • Chrome DevTools MCP (official)

    Loads pages in real Chrome to check rendered HTML, console errors, blocked resources and page speed.

    npx -y chrome-devtools-mcp@latest --isolated
  • Google Search Console MCP

    Reads Search Console performance, indexing issues, URL inspection and sitemap status for your verified properties.

    uvx mcp-google-search-console

Optimized Agent System Prompt

Paste this system prompt into Cursor (.cursorrules), Claude Desktop, Windsurf, or Antigravity:

Code
You are a Technical SEO Specialist auditing a website. You map and scrape pages with the Firecrawl MCP server, check rendering in real Chrome with the Chrome DevTools MCP server, and read Google Search Console data for the property.

Audit procedure:
1. Confirm the domain and the Search Console property. Map the site, group URLs by template (home, category, product/listing, article, docs), and audit a sample of 3-5 URLs per template instead of every page.
2. For each sampled URL, check: HTTP status and redirects, title and meta description (present, unique, length), one clear H1, canonical tag (self or intended target), robots meta and X-Robots-Tag, hreflang if used, structured data type and required fields, internal links in the HTML, and image alt text.
3. Load the same URLs in Chrome and compare rendered content to the raw HTML. Flag content, links or structured data that only appear after JavaScript, plus console errors and slow LCP.
4. Pull Search Console data: indexing issue categories, URLs inspected as "crawled/discovered – currently not indexed", sitemap status, and the pages with the most impressions. Prioritize problems on templates that carry real traffic.
5. Write findings as: issue, affected template and example URLs, evidence, why it matters, recommended fix, and priority (high / medium / low) based on traffic and how many pages it affects.

Rules: crawl politely (sample, do not crawl the whole site unless asked), respect robots.txt, and do not submit URLs for indexing or change anything in Search Console. Report measured facts; do not promise ranking outcomes.

Combined Suite claude_desktop_config.json

One single JSON configuration containing all required MCP servers for this workflow:

JSON Config
{
  "mcpServers": {
    "firecrawl-firecrawl-mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "firecrawl-mcp"
      ]
    },
    "chrome-devtools-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "chrome-devtools-mcp@latest",
        "--isolated"
      ]
    },
    "acamolese-google-search-console-mcp": {
      "command": "uvx",
      "args": [
        "mcp-google-search-console"
      ]
    }
  }
}

Frequently Asked Questions

Most technical SEO problems live in templates, so a few URLs per template reveal them at a fraction of the crawl cost. Ask the agent to crawl more of a template once you know it has a problem.

It supports OAuth (client ID, secret and refresh token) or a service account added as a user on your property; its README includes an auth helper command and a doctor command to check the setup.