MCP server

Give your coding agent a technical SEO audit

SECrawl's MCP server lets Claude Code, Cursor and any other MCP client read your site's audit, fix what it finds in your codebase, recheck the affected URLs and confirm each fix actually landed.

Start free See a sample audit

Audit, ask, fix, verify

Step 1

Audit

SECrawl crawls your site and runs 40+ technical SEO, AI-readiness and render checks.

Step 2

Ask

Your agent asks what is wrong and gets prioritised issues with the exact affected URLs.

Step 3

Fix

It fixes the templates in your codebase, then marks those issues resolved.

Step 4

Verify

It rechecks just those URLs. Each fix comes back verified_fixed — or regressed.

The last step is what makes this more than a report. Marking an issue resolved is a claim, not a result: a recheck of those URLs turns it into verified_fixed, and a fix that didn't land — or broke again — comes back as regressed instead of quietly dropping off the list.

Connect in a minute

1. Create a free account and run a crawl of your site.

2. Create an API key under Settings → API keys. It is shown once — copy it.

3. Add the server. In Claude Code:

claude mcp add --transport http secrawl \
  https://secrawl.com/api/mcp/ \
  --header "Authorization: Bearer sk_secrawl_..."

In Cursor, add it to ~/.cursor/mcp.json (or .cursor/mcp.json in a project) and set SECRAWL_API_KEY in your environment, so the key never lands in a file you might commit:

{
  "mcpServers": {
    "secrawl": {
      "url": "https://secrawl.com/api/mcp/",
      "headers": { "Authorization": "Bearer ${env:SECRAWL_API_KEY}" }
    }
  }
}

Any client that speaks streamable HTTP works the same way: the URL above plus a bearer token. Keep the trailing slash.

Things to ask it

"What are the three most important technical SEO problems on my site?"

"Which templates produce the missing meta descriptions? Fix them."

"I deployed the canonical fix — recheck those URLs and tell me what is still broken."

"What changed between last week's crawl and this one?"

Tools

Every tool works on every plan, including Free, except the three marked Professional, which power the verify step.

Tool	What it does
list_crawls	Your crawls, newest first.
get_crawl_health	Overall and per-category scores, and every open issue.
get_recommendations	Prioritised fixes, with platform-specific solutions.
get_issue_pages	The affected URLs for an issue, with status, redirects, indexability and URL-pattern clusters.
get_audit_findings	Cross-page findings: robots.txt, sitemaps, hreflang, canonicals, crawler parity.
get_render_evidence	Raw vs rendered HTML evidence for JavaScript-rendered pages.
describe_checks	What each check tests, its threshold, and why it matters.
list_supported_issue_types	Every issue key the audit can report.
get_crawl_stats	Pages indexed for a crawl.
set_issue_status	Mark issues acknowledged, false positive or resolved — by URL or pattern.
get_issue_status	The annotations recorded for a site.
recheck Professional	Re-fetch up to 50 URLs and compare every check before and after.
get_recheck Professional	Recheck results, with verified_fixed or regressed verdicts.
diff_crawls Professional	New and resolved issue URLs between two crawls of a site.

Point your agent at your site

Create a free accountNo credit card required.