How to Install seo-monster: Free Schema Checks and an "Is This Quotable by AI?" Score

Most "free SEO tools" are free the way a timeshare presentation is free.

seo-monster is the real thing. MIT licence, 70 tools, runs on your own computer with your own logins. We installed it on Oct 31, 2026, pointed it at live pages, and the very first check found a real problem.

Here's how to set it up and what to check first.

Quick words (one line each):

• MCP server: a plug-in that gives an AI assistant (Claude, Cursor, VS Code) new tools it can run for you.

• Schema (JSON-LD): a hidden block of code that tells Google and AI exactly what a page is about: business name, address, product, price.

• Search Console: Google's free dashboard that shows how your site does in Google search.

• GA4: Google Analytics 4, Google's free visitor-tracking dashboard.

• IndexNow: a free way to tell Bing "hey, this page just changed."

Step 1: Do a 60-second safety check

Never install a GitHub tool blind. Here's what seo-monster looked like when we checked it on Oct 31, 2026:

1. Licence: MIT. Free to use, including for client work.

2. Latest release: v0.9.3 on PyPI, released Aug 29, 2026. Last commit Sep 1, 2026.

3. Stars: 559, with 84 forks (as of Oct 31, 2026). People actually use this.

4. Runs locally. It uses your own Google login. Nothing goes to a middleman.

5. Costs: the core is free. The optional paid parts are DataForSEO, AI-engine API keys (only for the ai_citation_track tool) and Google Ads. Skip them and you pay nothing.

Step 2: Install it

seo-monster is a Python tool. The fastest route uses uv, a small Python helper.

1. Install uv if you don't have it (one command, from the uv website).

2. Run:

uvx seo-monster

3. Prefer clicking? The README has a one-click install link for Cursor and VS Code, plus a Claude Desktop bundle (a .mcpb file you double-click).

When we installed it, it listed 70 tools. That's your new toolbox.

Step 3: Connect Google (only if you want traffic data)

seo-monster setup

This connects your Search Console and GA4 with your own Google login. The page checks in Steps 4 and 5 work without this step, so you can try them first and connect Google later.

Step 4: Check your schema

Open Claude, Cursor or VS Code and ask:

Run validate_schema on https://yoursite.com/

It lists every schema block on the page and what's wrong with each one.

What we found: on Oct 31, 2026 we ran it on a live small-business homepage. Result: zero schema blocks. Not broken schema. No schema at all. Google and AI had to guess who the business was.

The fix is a basic Organization schema block with the business name, website, logo and contact details, using the exact same facts you show everywhere else. (Our brand facts post below covers which facts.)

Step 5: Score a page for AI quotes

This is the fun one. Ask:

Run ai_citation_readiness on https://yoursite.com/blog/your-best-post

It grades how easy the page is for an AI answer engine to quote. We ran it on one of our own blog posts on Oct 31, 2026:

1. Overall: 0.8 ("high").

2. Easy to pull out: 1.0.

3. Statistics: 1.0.

4. Cited sources: 1.0.

5. No keyword stuffing: 1.0.

6. Quotations: 0.0. Ouch.

So the post had numbers and sources, but not one real quote from a real person. Easy fix: add one quoted line from your source, with the name and date.

It also reports word count and the most-repeated word (ours: 1,634 words, top word at 2.9%), so you spot stuffing fast.

Step 6: Use the extras (carefully)

1. IndexNow pings: tell Bing a page changed, right after you publish. Free.

2. Before/after reports: see what changed in Search Console after you edited a page. Perfect for page refreshes.

3. Search Console and GA4 questions: "Which pages lost clicks this month?" in plain English.

Leave the Cloudflare write tools off. They only work if you switch on a special "destructive" setting. Don't. Some other tools also write things (sitemap submits, indexing requests), so only run those on sites you're allowed to change.

One thing to ignore

seo-monster's own notes say no engine reads llms.txt. Google said the same thing in its Oct 1, 2026 Search Central video: it doesn't use llms.txt. So if any tool gives you points for adding one, skip it. Spend that time on schema and quotes.

What it costs

Nothing for everything in this post. MIT licence, your own logins, a free PageSpeed key if you want speed data. The paid add-ons are optional.

Do this today

1. Run uvx seo-monster.

2. Ask your AI assistant to run validate_schema on your homepage.

3. If it finds zero blocks, add a basic Organization schema.

4. Run ai_citation_readiness on your best blog post.

5. If quotations score 0, add one real quote with a name and date.

Related reading

• Write pages AI can quote

• Build a brand facts page

• Install geoptimize to check pages before they go live

Sources

• GitHub: avansaber/seo-monster

• Google Search Central (Martin Splitt and John Mueller) - Do sitemaps still matter? (2026-10-01)

Don't worry if this seems like it's a pain in the ass for someone who doesn't build websites. That's because it is. The good news: just click the link below, pull the trigger, and we'll take care of the rest.