MCP.Market

CrazySEO MCP server

by bolidee·io.github.bolidee/crazyseo·v0.1.2·4 stars

Audit whether ChatGPT, Gemini and Perplexity recommend a site. Readiness checks need no keys.

B75/100grade B
What users say
No reviews yet
Be the first
Safety scan
B75/100

full report

Adoption
Growing

4 stars79 downloads/wk

Reviews

Write one

Nobody has reviewed CrazySEO yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

CrazySEO tools (4, 1 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

  • audit_visibility

    Measure whether AI assistants actually recommend a business. Infers what the site sells, generates the

  • citation_gap

    Show which third-party sources (directories, review sites, communities) AI engines cite on the buyer

  • engine_status

    List which AI engines this installation can query, based on which API keys are present in the environment. Never reveals key values.

  • readiness_checkwrite action

    Run 11 technical checks on a website that determine whether AI assistants can read and understand it:

Public scan report

scanner v0.1.9 · 2026-09-24 · same rubric, same numbers if you re-run it

no findings
  • Code scan15 source files scanned25/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitystatic API keys via environment variables6/15
  • Maintenancelast push 50 days ago12/15
  • Maintainer identityregistry namespace matches repository owner6/10
Overall 75/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

How CrazySEO compares

Against every graded Web search MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score
75/100
higher than 45% of 3,472 Web search servers
Tools
4
the typical Web search server has 5
Tools that change things
25%
typical: 0%. Fewer write tools means less an agent can break
GitHub stars
4
more than 75% of 1,783 Web search servers
Downloads a week
79
more than 52% of 816 Web search servers with a package

What its tools let an agent do: 3 read · 1 change

Install CrazySEO in Claude Code, Cursor or VS Code

Runs npx -y crazyseo on your machine. Read the scan report first; the gateway never runs local packages.

claude mcp add crazyseo -- npx -y crazyseo
Add to Cursor

What the publisher says

From the CrazySEO repository's README, as published. We do not edit it. Read it on GitHub

CrazySEO

Find out whether ChatGPT, Gemini and Perplexity recommend your site — and why they don't.

An open-source GEO / AEO auditor and MCP server. No account. No card. Your keys.

Your customers have stopped googling. They ask an assistant "best CRM for small law firms?" and buy whatever it names. If it doesn't name you, you don't exist — and no analytics tool will tell you, because there is no click to measure.

This audits it directly: it asks the questions your buyers ask, records the answers word for word, and shows you which sources the models cite instead of you.

Why another one

Most tools in this space ask the model "what do you think of Acme?" — which tells you nothing, because naming the brand in the prompt guarantees it shows up in the answer.

This asks blind. Your brand never appears in the question, so the answer you get back is the answer a real buyer gets. That is the whole difference between a number you can act on and a number that flatters you.

Three more things a prompt wrapper can't do:

raw answer, so you can paste it into ChatGPT yourself and reproduce it.

  • Verbatim receipts. Every result carries the exact prompt and the model's

never quietly drops an engine and calls the remaining average your score. If a site can't be reached, it refuses to score it rather than inventing a number.

  • Honest coverage. If an engine fails or rate-limits, the report says so. It
  • A real readiness audit. 11 checks against your actual site — not a prompt.

Quick start — no API keys needed

The readiness audit runs against any site with zero configuration and zero cost:

npx crazyseo readiness yourbrand.com

Nothing to install, no signup, no key. Or run it from source:

git clone https://github.com/d08m/crazyseo && cd crazyseo
npm install && npm run build
node dist/cli.js readiness yourbrand.com
AI-READINESS  stripe.com                                    82/100

  ✓ Served over HTTPS
  ✓ AI crawlers allowed
  ✓ Structured data (schema.org)
  ✓ llms.txt published (bonus)
  ✓ Title + meta description
  ✓ Readable without JavaScript
  ✓ Canonical URL set
  ✓ Social preview tags (Open Graph)
  ✗ Single, clear H1
      2 H1 tags — ambiguous which is the main topic
  ✗ Images have alt text
      27 of 31 images missing alt text
  ✓ Sitemap referenced
  ✓ No broken internal links

  → 2 issues to fix

(That's real output — stripe.com, unmodified.)

11 scored checks, all first-party fetches: HTTPS · AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended…) unblocked in robots.txt · schema.org markup · title + meta · readable without JavaScript · canonical · Open Graph · single-H1 structure · image alt-text coverage · sitemap referenced · broken internal links (sampled).

Plus llms.txt, shown but deliberately unscored: Google's own AI-optimization guidance says it has no effect in Google Search, so it can't fairly cost anyone points — but some agents do read it, so it's still worth seeing.

Exits non-zero when checks fail, so you can gate a deploy on it:

- run: npx crazyseo readiness ${{ env.SITE }}

Machine-readable output for pipelines:

npx crazyseo readiness yourbrand.com --json > report.json

The visibility audit — bring your own keys

To measure what the engines actually say, add whichever keys you have. Any one is enough; more engines means more coverage.

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

CrazySEO: common questions

Is CrazySEO MCP server safe?
Mostly: it is graded B (75/100). Read the CrazySEO safety report
How do I install CrazySEO?
It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does CrazySEO need an API key?
Yes. The registry entry asks for OPENAI_API_KEY, GOOGLE_API_KEY, PERPLEXITY_API_KEY.
Is CrazySEO maintained?
The last commit was 89 days ago (2026-08-05). The latest release is v0.1.2.
What can I use instead of CrazySEO?
Servers from other publishers that do the same job: Perplexity API Platform (moved to ai.perplexity/mcp-server), Perplexity API Platform MCP server and Tolgee MCP server.
How does CrazySEO compare with other Web search MCP servers?
safety score 75/100 (higher than 45% of 3,472 Web search servers); tools 4 (the typical Web search server has 5); tools that change things 25% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.

Alternatives to CrazySEO

Same job from other publishers: the closest match first, then the best rated.

  • Perplexity API Platform (moved to ai.perplexity/mcp-server)
    Real-time web search, reasoning, and research through Perplexity's API
    A
  • Perplexity API Platform
    Real-time web search, reasoning, and research through Perplexity's API
    A
  • Tolgee
    Your app's translations in Tolgee: search keys, create translations, trigger machine translation
    A
  • GEO Tracker by DigestSEO
    Track brand citations across seven AI search surfaces. Free OSS; optional EUR 99 client-ready audit.
    B
  • HIG Doctor
    Apple Human Interface Guidelines search, lookup, and project audits for AI coding agents.
    B