The scanner checks 7 things: 1. robots.txt - are you blocking AI crawlers? 2. Structured data - JSON-LD that LLMs can parse 3. Metadata - title, description, OG tags 4. Content structure - headings, lists, semantic HTML 5. llms.txt - the emerging standard (like robots.txt for AI) 6. Citation signals - author, date, canonical URL 7. Extractability - FAQs, tables, definition lists
Tech stack: FastAPI + Jinja2 + Tailwind (no React). Hosted on Cloud Run. 181 tests.
The thesis: SEO optimized for Google's index. GEO (Generative Engine Optimization) optimizes for AI training data and retrieval. Different game, different rules.
Happy to discuss the technical approach or GEO strategy. Try it on your own site and let me know what you find.