@ihmissuti/geo-tools
v0.1.0
Published
MCP server for Generative Engine Optimization (GEO) - AI search visibility tools by Kimmo Ihanus
Maintainers
Readme
@ihmissuti/geo-tools
MCP server for Generative Engine Optimization (GEO) - AI search visibility tools.
Connect Claude Desktop, Cursor, or any MCP-compatible client to audit websites for AI visibility, generate optimized schemas, and evaluate devtools for the agent era.
Quick Start • Available Tools • About GEO
What is this?
GEO (Generative Engine Optimization) is the practice of optimizing content and technical infrastructure so AI systems like ChatGPT, Perplexity, and Claude can find, understand, and recommend your products.
This MCP server provides tools to:
- 📊 Audit webpages for AI search visibility
- 🏷️ Generate schemas optimized for LLM understanding
- 🔧 Score devtools for AI agent compatibility
- 📋 Get recommendations to improve AI visibility
Quick Start
Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"geo-tools": {
"command": "npx",
"args": ["-y", "@ihmissuti/geo-tools"]
}
}
}Restart Claude Desktop and look for the hammer icon (🔨).
Cursor
Add to .cursor/mcp.json:
{
"mcpServers": {
"geo-tools": {
"command": "npx",
"args": ["-y", "@ihmissuti/geo-tools"]
}
}
}Global Installation
npm install -g @ihmissuti/geo-toolsThen configure:
{
"mcpServers": {
"geo-tools": {
"command": "geo-tools"
}
}
}Available Tools
webpage_geo_audit
Audit a webpage for AI search visibility. Analyzes schema markup, content structure, and technical factors.
Audit https://example.com for AI visibilityReturns:
- Overall GEO score (0-100)
- Grade (A-F)
- Strengths and gaps analysis
- Technical factors (schema, headings, code blocks, etc.)
generate_schema
Generate optimized Schema.org markup for AI visibility.
Generate organization schema for my company "Acme Inc" at https://acme.comSupported types:
organization- Company/brand schemasoftware- SoftwareApplication for devtoolsperson- Personal brand schemafaq- FAQPage for common questionshowto- HowTo for tutorials/guides
agent_friendly_score
Evaluate a developer tool for AI agent compatibility.
Score Resend for agent-friendlinessEvaluates:
- SDK & API design (30 pts)
- Documentation quality (25 pts)
- Training data presence (20 pts)
- MCP integration (15 pts)
- Time to working (10 pts)
geo_recommendations
Get prioritized recommendations based on an audit.
Give me GEO recommendations based on the audit resultsReturns actionable improvements sorted by impact.
Example Queries
Full GEO Audit Flow
1. Audit https://myproduct.com for AI visibility
2. Generate organization and software schemas for my product
3. What are the top 3 things I should fix to improve my AI search presence?DevTool Evaluation
Compare the agent-friendliness of Resend vs SendGridSchema Generation
Generate FAQ schema for these questions:
- How do I get started?
- What's the pricing?
- How does authentication work?About GEO
GEO is bigger than "SEO for ChatGPT." It's about understanding that AI assistants now make infrastructure decisions for developers.
When a developer asks Cursor to "add email functionality," the AI:
- Picks the service
- Writes the integration
- Runs the install command
The tools that AI can understand and work with get recommended. The tools it can't work with become invisible.
The new funnel:
- ~~Marketing → Landing → Docs → Trial → Conversion~~
- Problem → AI suggestion →
npm install→ Subscription
Learn more:
Author
Kimmo Ihanus
- Website: kimmoihanus.com
- GitHub: @kimmoihanus
- Twitter: @kimmoihanus
License
MIT
