免费扫描
MCP + API

GEOScore for AI Agents

Add AI search readiness scanning to Claude Code, Cursor, or any MCP-compatible tool. One line of config — your AI agent can scan any website and get structured GEO results.

Quick Start: MCP Server

Add this to your project's .mcp.json (or ~/.claude.json for global access):

.mcp.jsonOne-line install
{
  "mcpServers": {
    "geoscore": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "geoscore-mcp"]
    }
  }
}

That's it. Your AI agent now has 3 GEO tools available.

Available MCP Tools

geo_scan

Scan a domain for AI search readiness. Returns a score (0-100), grade (A-F), and 11 technical checks with specific issues and fix suggestions.

Try it:"Scan example.com for GEO readiness"
geo_get_profile

Get a domain's GEO score history — latest score, best score, scan count, and dates. Track improvement over time.

Try it:"Show GEO profile for example.com"
geo_compare

Compare two domains side by side with score tables and per-metric breakdown. Perfect for competitive analysis.

Try it:"Compare example.com vs competitor.com"

Example: What It Looks Like

You

Scan my site geoscoreai.com and tell me what to fix for AI search

Claude

I'll scan your site using GEOScore.

GEO Score: 82/100 (Grade: A-)

[PASS] robots.txt (9/10)
[PASS] Structured Data (8/10)
[WARN] llms.txt (6/10) — missing /.well-known/llms.txt
[WARN] Meta Tags (7/10) — 3 pages missing og:description

Priority fixes:
1. Add /.well-known/llms.txt (copy from /llms.txt)
2. Add og:description to /checks, /for, /blog pages

Works With

Claude Code

Add to .mcp.json — scan sites while developing, fix GEO issues in the same session.

Cursor

Add to MCP settings — let Cursor's AI check your site's GEO readiness before deployment.

ChatGPT / GPT Actions

Use the REST API to build a custom GPT that answers "Is my site ready for AI search?"

CI/CD Pipelines

Add GEO score checks to your deployment pipeline. Catch AI search regressions before they go live.

REST API Reference

Prefer the MCP server for AI agent integration. Use the REST API for custom integrations, dashboards, or automation scripts.

RequestPOST
POST https://api.geoscoreai.com/api/v1/geo/check
Content-Type: application/json

{
  "domain": "example.com"
}
Response200 OK
{
  "score": 72,
  "grade": "B+",
  "checks": {
    "robots_txt": { "score": 9, "status": "pass", "details": "..." },
    "llms_txt": { "score": 0, "status": "fail", "details": "..." },
    "structured_data": { "score": 8, "status": "pass", "details": "..." },
    "meta_tags": { "score": 7, "status": "warn", "details": "..." },
    ...
  },
  "recommendations": [
    "Create a /llms.txt file to guide AI models",
    "Add FAQ schema to your product pages",
    ...
  ]
}
MethodEndpointDescription
POST/api/v1/geo/checkRun a standard GEO scan on a domain
POST/api/v1/geo/check-deepRun a deep scan with multi-page analysis
GET/api/v1/geo/site/:domainGet a domain's profile and score history
GET/api/v1/geo/site/:d1/vs/:d2Compare two domains side by side
GET/api/v1/geo/sites/recentList recently scanned domains

Start Using GEOScore

Install the MCP server and let your AI agent scan websites, or try a free scan now.

Run a Free Scan