GeoSEO

Generate GEO assets for AI discoverability — llms.txt, robots.txt, sitemap.xml, and JSON-LD.

GeoSEO is a framework-agnostic CLI and Node.js library that auto-generates the files AI crawlers look for — llms.txt, robots.txt with AI crawler rules, sitemap.xml, and JSON-LD structured data. Define your pages in one config file, generate everything at build time. Zero runtime overhead.

Installation

# Install globally
npm install -g @glincker/geo-seo

# Or use with npx (no installation)
npx @glincker/geo-seo generate

Quick Start

# Create a starter config file
geo-seo init

# Edit geo-seo.config.ts with your site info, then:
geo-seo generate

# Generate only llms.txt
geo-seo generate --only llms-txt

# Output to a custom directory
geo-seo generate --out ./dist

Configuration

Run geo-seo init to create a starter geo-seo.config.ts:

import { defineConfig } from '@glincker/geo-seo';

export default defineConfig({
  site: {
    name: 'My Website',
    url: 'https://example.com',
    description: 'A great website',
    logo: '/logo.png',
  },
  organization: {
    name: 'My Company',
    url: 'https://example.com',
    logo: 'https://example.com/logo.png',
    sameAs: [
      'https://twitter.com/mycompany',
      'https://github.com/mycompany',
    ],
  },
  pages: [
    { path: '/', title: 'Home', description: 'Welcome to our site', changefreq: 'daily', priority: 1.0 },
    { path: '/about', title: 'About', description: 'About us' },
    { path: '/blog', title: 'Blog', description: 'Latest posts', changefreq: 'weekly' },
    { path: '/pricing', title: 'Pricing', description: 'Plans and pricing', schemaType: 'Product' },
    { path: '/faq', title: 'FAQ', description: 'Frequently asked questions', schemaType: 'FAQPage', schemaProps: {
      questions: [
        { question: 'What is GEO?', answer: 'Generative Engine Optimization...' },
      ],
    }},
  ],
  robots: {
    aiCrawlers: 'allow', // 'allow' | 'block' | 'selective'
  },
  output: './public',
});

Generated Files

llms.txt

Following the llmstxt.org specification:

# My Website

> A great website

- [Home](https://example.com/): Welcome to our site

## Pages

- [About](https://example.com/about): About us
- [Blog](https://example.com/blog): Latest posts
- [Pricing](https://example.com/pricing): Plans and pricing
- [FAQ](https://example.com/faq): Frequently asked questions

robots.txt

With AI crawler rules:

User-agent: *
Allow: /

# AI Crawlers: Allowed
# All AI crawlers can access this site

Sitemap: https://example.com/sitemap.xml

Supports 16 AI crawlers: GPTBot, ChatGPT-User, ClaudeBot, Claude-Web, PerplexityBot, Amazonbot, CCBot, Google-Extended, FacebookBot, Bytespider, Applebot-Extended, cohere-ai, Diffbot, ImagesiftBot, Omgilibot, YouBot.

sitemap.xml

Standard sitemap with optional changefreq and priority:

<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
  <url>
    <loc>https://example.com</loc>
    <changefreq>daily</changefreq>
    <priority>1.0</priority>
  </url>
  <url>
    <loc>https://example.com/about</loc>
  </url>
</urlset>

JSON-LD

Generate structured data for any page:

import { generateJsonLd } from '@glincker/geo-seo';

const orgSchema = generateJsonLd('Organization', config);
// { "@context": "https://schema.org", "@type": "Organization", ... }

const faqSchema = generateJsonLd('FAQPage', config, faqPage);
// { "@context": "https://schema.org", "@type": "FAQPage", "mainEntity": [...] }

Supported types: Organization, WebSite, WebPage, Article, BlogPosting, FAQPage, Product, BreadcrumbList.

AI Crawler Modes

Control which AI crawlers can access your site:

// Allow all AI crawlers (default)
robots: { aiCrawlers: 'allow' }

// Block all AI crawlers
robots: { aiCrawlers: 'block' }

// Selective: allow some, block others
robots: {
  aiCrawlers: 'selective',
  allow: ['GPTBot', 'ClaudeBot', 'PerplexityBot'],
  block: ['Bytespider', 'CCBot'],
}

You can also add custom rules and crawl delays:

robots: {
  aiCrawlers: 'allow',
  crawlDelay: 2,
  customRules: [
    { userAgent: 'BadBot', disallow: ['/'] },
  ],
}

CLI Commands

| Command | Description | |---------|-------------| | geo-seo generate | Generate all GEO assets | | geo-seo init | Create a starter config file | | geo-seo validate | Validate your config file |

CLI Flags

| Flag | Short | Description | |------|-------|-------------| | --out <dir> | -o | Output directory (default: ./public) | | --only <type> | | Generate only: llms-txt, robots-txt, sitemap | | --config <path> | -c | Path to config file (default: geo-seo.config.ts) | | --help | -h | Show help message | | --version | -v | Show version |

Programmatic API

Use GeoSEO in your build pipeline:

import {
  defineConfig,
  generateAll,
  generateLlmsTxt,
  generateRobotsTxt,
  generateSitemap,
  generateJsonLd,
  validateConfig,
} from '@glincker/geo-seo';

// Define config with type safety
const config = defineConfig({
  site: { name: 'My Site', url: 'https://example.com', description: '...' },
  pages: [{ path: '/', title: 'Home', description: '...' }],
});

// Validate before generating
const errors = validateConfig(config);
if (errors.length > 0) {
  console.error('Config errors:', errors);
  process.exit(1);
}

// Generate all files to output directory
const result = await generateAll(config);
console.log(`Generated ${result.count} files`);

// Or generate individual assets as strings
const llmsTxt = generateLlmsTxt(config);
const robotsTxt = generateRobotsTxt(config);
const sitemap = generateSitemap(config);
const jsonLd = generateJsonLd('Organization', config);

Config File Formats

GeoSEO looks for config files in this order:

1. geo-seo.config.ts
2. geo-seo.config.js
3. geo-seo.config.mjs
4. geo-seo.config.json

Or specify a custom path with --config:

geo-seo generate --config ./my-config.ts

CI/CD Integration

Run GeoSEO as part of your build:

# .github/workflows/build.yml
- name: Generate GEO assets
  run: npx @glincker/geo-seo generate --out ./public

- name: Verify GEO score
  run: npx @glincker/geo-audit https://yourdomain.com --fail-under 80

Framework Adapters

GeoSEO integrates with popular frameworks via subpath exports — no extra packages needed:

Next.js

// next.config.ts
import { withGeoSEO } from '@glincker/geo-seo/next';
export default withGeoSEO(geoConfig)({ reactStrictMode: true });

Vite (Vue / Svelte / Remix / SolidStart)

// vite.config.ts
import { geoSeoPlugin } from '@glincker/geo-seo/vite';
export default defineConfig({ plugins: [geoSeoPlugin(geoConfig)] });

Astro

// astro.config.mjs
import { geoSeoIntegration } from '@glincker/geo-seo/astro';
export default defineConfig({ integrations: [geoSeoIntegration(geoConfig)] });

Nuxt

// nuxt.config.ts
import { geoSeoModule } from '@glincker/geo-seo/nuxt';
export default defineNuxtConfig({ modules: [geoSeoModule(geoConfig)] });

All adapters also export getJsonLdScript() for page-level JSON-LD generation:

import { getJsonLdScript } from '@glincker/geo-seo/vite'; // or /next, /astro, /nuxt
const jsonLd = getJsonLdScript('WebPage', geoConfig, page);
// Use in <script type="application/ld+json">{JSON.stringify(jsonLd)}</script>

How It Connects to GeoKit

GeoSEO is one part of the GeoKit toolkit:

1. geo-audit — Audit your site's AI-readiness (score 0-100)
2. geo-seo — Generate the files that improve your score
3. geomark — Convert URLs to clean markdown for LLM context

Typical workflow:

Run geo-audit -> Score 45/100
  -> Install geo-seo -> Generate llms.txt + robots.txt + sitemap
  -> Run geo-audit again -> Score 85/100

Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines.

License

MIT License - see LICENSE for details.

Links

• Website: geo.glincker.com
• GitHub: github.com/glincker/geokit
• Issues: github.com/glincker/geokit/issues
• npm: npmjs.com/package/@glincker/geo-seo

Made by Glincker | geo.glincker.com