Skip to main content
MCP server CX Codex

seo-audit-console — Claude Code MCP server: SEO audit

Joins Search Console history with a live crawl, then ranks each finding by the clicks a fix could bring back.

Install into Claude Code · Platform CLI All methods and platforms ↓
claude mcp add --env GOOGLE_APPLICATION_CREDENTIALS=<GOOGLE_APPLICATION_CREDENTIALS> --env SAC_DATA_DIR=<SAC_DATA_DIR> --env DATAFORSEO_USERNAME=<DATAFORSEO_USERNAME> --env DATAFORSEO_PASSWORD=<DATAFORSEO_PASSWORD> --env DATAFORSEO_CACHE_DAYS=<DATAFORSEO_CACHE_DAYS> --transport stdio seo-audit-console -- npx -y @houtini/seo-audit-console
Open on GitHub

What seo-audit-console does

The server runs inside Claude and pulls two things together: your Google Search Console history and a fresh crawl of the site. Findings come out ordered by the clicks a fix could recover, not by how alarming the check name sounds. Each one arrives with the fix written out — a redirect rule, JSON-LD, internal links.

The core audit needs only GOOGLE_APPLICATION_CREDENTIALS, the JSON key of a Google Cloud service account. That account has to be added as a user on your Search Console property, or the server sees nothing. DataForSEO, Majestic and CRUX_API_KEY come in later, once your questions turn to competitors and links. An empty value switches that tier off, and placeholder text such as your-key-here counts as empty too — the server says so in its log, so a tier that stays off is worth a look there.

Node.js 20 or newer is required. Your data stays on your machine. Three ways to read the result: the dashboard drawn in the chat, the full page in a browser at a local link like http://127.0.0.1:53817/dashboard?siteUrl=…, or one self-contained HTML file to send a client. The link answers on 127.0.0.1 only, and the port changes on every server start, so use the link from the latest reply.

What it helps with

  • find pages losing clicks at a healthy position
  • untangle keyword cannibalisation between pages
  • check whether a robots-blocked section still earns traffic
  • export a client-ready report as a single HTML file
  • compare the current crawl against the previous one

Example prompts for the agent

  • Run an SEO audit on mysite.com
  • Refresh mysite.com
  • Generate the fix for finding #1

How to install seo-audit-console

Platform CLI

claude mcp add --env GOOGLE_APPLICATION_CREDENTIALS=<GOOGLE_APPLICATION_CREDENTIALS> --env SAC_DATA_DIR=<SAC_DATA_DIR> --env DATAFORSEO_USERNAME=<DATAFORSEO_USERNAME> --env DATAFORSEO_PASSWORD=<DATAFORSEO_PASSWORD> --env DATAFORSEO_CACHE_DAYS=<DATAFORSEO_CACHE_DAYS> --transport stdio seo-audit-console -- npx -y @houtini/seo-audit-console

Replace <…> with your own values. 3 more environment variables are in the server README

Step 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 seo-audit-console

How do I install seo-audit-console in Claude Code?

With one command: `claude mcp add seo-audit-console -e GOOGLE_APPLICATION_CREDENTIALS=<GOOGLE_APPLICATION_CREDENTIALS> -e DATAFORSEO_PASSWORD=<DATAFORSEO_PASSWORD> -e MAJESTIC_API_KEY=<MAJESTIC_API_KEY> -e CRUX_API_KEY=<CRUX_API_KEY> -- npx -y @houtini/seo-audit-console`. The service account key is required; the other variables can stay empty.

What does seo-audit-console do?

It reads your Google Search Console history, crawls the site, and ties every finding to the traffic at stake. Each problem comes with a suggested fix.

What do I need besides the service account key?

Node.js 20 or newer. The service account from the JSON key also has to be added as a user on your Search Console property, otherwise the server cannot see your data.

Similar MCP servers

Everything for SEO and GEO →

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor