Glippy GEO

By JAN WILLEM BOBBINK

Overview

Glippy is a website audit tool for marketers, SEO specialists and web developers. It checks how AI-ready and LLM-friendly a website is: can AI search engines and AI agents find it, read it, and cite it? Use it to optimize your website for AI search, also called AI SEO or Generative Engine Optimization (GEO). Name a website and Glippy returns an interactive report: - An overall GEO score and grade - 16 scored categories, including structured data, semantic HTML, machine readability, entity and authority signals, citability and content freshness - Which AI crawlers robots.txt blocks or allows - Whether the site publishes a sitemap and llms.txt - A prioritised list of fixes you can copy, or ask for drafts of You can also compare up to eight sites side by side, for example your site against competitors, and see who leads in each category. How Glippy fetches: it reads only public files of the site you name (the homepage, robots.txt, sitemap.xml, llms.txt and standard /.well-known discovery files), identifies itself as GlippyBot, and does not fetch a site whose robots.txt disallows GlippyBot. It analyses the homepage, not the whole site. It does not sign in anywhere, change anything on the site, or need an account.

Tools

analyze_domain

ChatGPT

Run a full GEO (Generative Engine Optimization) analysis of a public website the user names: 16 scored categories, robots.txt / llms.txt / sitemap discovery, AI-crawler rules, and prioritised recommendations. Analyses the homepage plus standard discovery files, not every page. Skips sites whose robots.txt disallows GlippyBot. Before analysing, decide what kind of page the site's homepage is (from what you know of the site or what the user said) and pass page_type: usually "homepage"; a news front or blog index is "index", a one-page local business site is "service", a single-product site is "product". Without it, checks that depend on the page type are shown but not scored. Decide before the first call: a re-run with another page_type is a new analysis and counts toward the per-site limit. Renders an interactive report widget.

check_llms_txt

ChatGPT

Check whether a domain publishes an llms.txt file (one request) and return its contents. Note: llms.txt is an emerging proposal not yet consumed by major AI models. The full analysis gives a small credit in Machine Readability when one exists; a missing one is listed as information, not as a fix to make.

check_robots_txt

ChatGPT

Read a domain's robots.txt (one request) and report which AI crawlers (training, search and user-triggered fetchers) it blocks or allows. This reports the stated policy only; it does not test what the site actually serves each crawler.

compare_domains

ChatGPT

Analyse and rank 2-8 public websites the user names, side by side: overall scores, a per-category comparison matrix, discovery signals, and the top issues per site. Each site's homepage is analysed under the same rules as analyze_domain. Before analysing, decide what kind of page the site's homepage is (from what you know of the site or what the user said) and pass page_type: usually "homepage"; a news front or blog index is "index", a one-page local business site is "service", a single-product site is "product". Without it, checks that depend on the page type are shown but not scored. Decide before the first call: a re-run with another page_type is a new analysis and counts toward the per-site limit. Renders an interactive comparison widget.

get_geo_summary

ChatGPT

Get a concise GEO readiness summary for a public website the user names: overall score, grade, top strengths, and the top issues to fix. Renders the report widget in summary form. Use analyze_domain for the full 16-category breakdown. Before analysing, decide what kind of page the site's homepage is (from what you know of the site or what the user said) and pass page_type: usually "homepage"; a news front or blog index is "index", a one-page local business site is "service", a single-product site is "product". Without it, checks that depend on the page type are shown but not scored. Decide before the first call: a re-run with another page_type is a new analysis and counts toward the per-site limit.