Crawlinx — Claude Code MCP server: what it does, how to install
Crawlinx — a hosted MCP server for automated technical SEO audits. It runs 156 checks, generates compact reports for AI agents, and compares scans after fixes.
no license 372 installs updated 2026-07-11
claude mcp add --transport http crawlinx https://crawlinx.run.toolsWhat Crawlinx does
Crawlinx runs as a remote service over Streamable HTTP without local binaries, account registration, or access tokens. An agent calls scan_url to inspect a domain, crawling 50 pages by default or up to 200 when setting max_pages. The engine inspects 156 technical points, covering title tags, canonical links, noindex flags, orphan URLs, internal PageRank, JS rendering differences, asset sizing, and sitemaps.
Output comes as structured JSON via get_report or as a concise table through get_report_markdown, designed to preserve prompt context. When an agent tackles a specific error code, get_issue_detail returns every affected URL alongside remediation advice.
After applying changes, the agent calls diff_scan to verify resolved errors and flag any regressions against the previous crawl. Because processing happens on the hosted endpoint, the host machine requires only outbound internet connectivity.
What it helps with
- detect broken canonical tags and orphan URLs across crawled pages
- spot layout or content discrepancies caused by client JavaScript rendering
- pull step-by-step remediation instructions for technical SEO defects
- compare sequential domain crawls to confirm fixes and catch regressions
Example prompts for the agent
- Start a technical SEO crawl on the target site with
scan_url. - Fetch the summary using
get_report_markdownand prioritize high-severity items. - Run
diff_scanagainst the prior scan to confirm our changes worked.
How to install Crawlinx
Platform CLI: remote server
claude mcp add --transport http crawlinx https://crawlinx.run.toolsPlatform CLI: remote server
codex mcp add crawlinx --url https://crawlinx.run.toolsConfig file
{
"mcpServers": {
"crawlinx": {
"url": "https://crawlinx.run.tools"
}
}
}Config file
{
"servers": {
"crawlinx": {
"type": "http",
"url": "https://crawlinx.run.tools"
}
}
}Platform CLI: remote server
gemini mcp add -t http crawlinx https://crawlinx.run.toolsStep by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Crawlinx
How to install Crawlinx in Claude Code?
Run claude mcp add --transport http crawlinx https://crawlinx.run.tools in your terminal.
Does Crawlinx require an API key or account?
No credentials or accounts are needed. The server communicates directly over HTTP via its remote endpoint.
What is the page limit for a single crawl?
Crawlinx scans 50 pages by default, but you can expand the crawl boundary up to 200 pages using max_pages.