Are you letting
the bots in?

ChatGPT uses OAI-SearchBot to index content for real-time citations. This is a different crawler from GPTBot (which is for training). If your robots.txt blocks OAI-SearchBot — or has a wildcard rule that catches it — ChatGPT cannot cite your site, regardless of how relevant your content is. Paste your URL above to check in under 10 seconds.

No — not directly. GPTBot is OpenAI's training crawler. Blocking it limits what GPT models learn during pre-training, but has no effect on real-time citations. The citation pipeline uses OAI-SearchBot, which is a separate user agent. Many sites that blocked GPTBot (a defensible choice) accidentally also blocked OAI-SearchBot because they used wildcard disallow rules. We check both.

The most common causes, in order of frequency:

• Blocked crawler. Your robots.txt has a wildcard or per-bot rule blocking the AI retrieval crawler.
• JavaScript-only content. Your content loads via client-side JavaScript. Most AI crawlers fetch raw HTML only.
• Missing entity signals. No structured data, no Wikidata presence, no named authorship — AI systems treat you as an anonymous source.
• noindex tag. A meta robots noindex directive on pages that should be publicly citable.

We check all four as part of the free scan.

GEO (Generative Engine Optimization) is the practice of making sure AI search engines can find, read, and cite your content. It's related to SEO but distinct in important ways. Traditional SEO asks: can Google rank this page? GEO asks: can AI systems extract a citable answer from it? The technical requirements differ — AI crawlers often don't render JavaScript, entity authority matters more than backlinks, and content structure for quotability is different from content structure for dwell time. Our GEO guide walks through all six signals.

To be citable by each major platform, you need to allow these retrieval crawlers:

• OAI-SearchBot — ChatGPT citations
• Claude-SearchBot — Claude citations
• PerplexityBot — Perplexity citations
• Googlebot — Gemini / Google AI Overviews
• Bingbot — Microsoft Copilot

The training crawlers (GPTBot, ClaudeBot, Google-Extended) are separate. Blocking those is fine; it only affects model training, not citations. See our full robots.txt guide.

llms.txt is a plain-text file at yourdomain.com/llms.txt that gives AI systems a structured summary of your site: what it does, who runs it, and what the key pages are. It was proposed in 2024 and is increasingly checked by Perplexity and other AI engines during indexing. Creating one takes under 20 minutes and contributes to your freshness and hygiene score. It won't fix access or readability problems on its own, but it's worth having once the bigger issues are resolved. See our llms.txt guide.

No. letthebots.in is stateless by design. Scans run fresh every time and no site data is persisted between sessions. If you enter your email to unlock the full fix guide, your email and scan score are stored in our email platform (Kit), but the scan results themselves are not. We run 6 live checks every time you scan — nothing is cached or stored about the site.

A score of 85 or above earns the Citation-Ready badge. It means your site passes all critical GEO signals: AI crawlers can reach you, your content is readable without JavaScript rendering, you have JSON-LD structured data covering key types, your entity is identifiable (sameAs links, authorship), your content is structured for extraction, and your date signals are current. Sites at this level are well-positioned to be cited when AI engines encounter relevant queries.