Frequently Asked Questions

Your AI Exposure Score (0-100) measures how visible and understandable your product is to AI systems like ChatGPT, Claude, Perplexity, and Gemini. It's calculated across 6 categories: AI Crawl Access, Content Quality, Product Clarity, Structured Data & Meta, Agent Readiness, and Trust & Social Proof — 25+ signals total.

AI assistants are increasingly used for product discovery. When someone asks ChatGPT or Perplexity 'what's the best tool for X', AI systems recommend products they can actually understand and access. A low score means AI can't properly crawl, parse, or recommend your product. A high score means you're visible and citable by AI systems — essentially AI-native SEO.

Our security scanner runs 19 passive, read-only checks on your website. It checks HTTPS configuration, security headers (CSP, X-Frame-Options, HSTS, etc.), exposed files (.env, .git, source maps), API key leaks in HTML and JavaScript bundles, admin endpoint exposure, CORS configuration, and WAF detection. Scans complete in under 15 seconds and are completely safe — no intrusive probing.

Yes. We fetch all JavaScript bundle files linked from your homepage and scan them against 17 known API key patterns — including Stripe secret keys, OpenAI keys, Anthropic/Claude keys, Google Gemini keys, Supabase service role keys, Razorpay keys, AWS access keys, and more. This is a common mistake in vibe-coded and AI-assisted apps where secrets accidentally get bundled into client-side code.

llms.txt is a plain-text file placed at /llms.txt on your website that gives AI crawlers a structured, human-readable summary of your product. Similar to how robots.txt guides general crawlers and sitemap.xml guides search engines, llms.txt helps LLMs like Claude, ChatGPT, and Perplexity understand your product quickly and accurately. We generate one for you based on your site scan.

Yes, completely free — no credit card required. Your first scan of any URL gives you an AI Exposure Score (0-100), a Security Grade (A-F), and your top 3 issues. You can scan any number of different URLs for free. Paid plans unlock full issue breakdowns, fix prompts, re-scanning, llms.txt generation, and more.

Starter ($19/month or $15/month billed annually) gives you 25 AI visibility scans and 25 security scans per month, full reports, all fix prompts, and llms.txt/JSON-LD generation. Pro ($49/month or $39/month billed annually) gives you unlimited scans, priority processing, and all upcoming features like weekly re-scan emails, competitor comparison, and score change tracking.

Fix prompts are copy-paste prompts tailored for Claude, Cursor, ChatGPT, and Gemini. Each issue in your report comes with a specific prompt you can paste directly into your AI coding tool to fix that issue. Instead of guessing what to tell Claude, you get the exact prompt that explains your site's problem and how to resolve it.

Completely safe. All scans are passive and read-only — we fetch public pages and check publicly accessible files and headers. We don't send login attempts, form submissions, fuzzing requests, or anything that could affect your site's performance or trigger alerts. Think of it as a sophisticated version of what a browser does when it visits your site.

The biggest wins come from: (1) adding a sitemap.xml, (2) creating a /llms.txt file with product information, (3) adding JSON-LD structured data (SoftwareApplication or Product schema), (4) writing a clear H1 headline and meta description, (5) adding social proof and quantifiable metrics to your homepage, and (6) ensuring your content is text-rich rather than JS-rendered. Our audit shows you exactly what to fix, scored by impact.

Yes, scan results are stored so you can access your report later via a unique URL. We do not store your website's private data — only the publicly visible information we check during the scan. See our Privacy Policy for full details.

You can scan any publicly accessible website. The scanner only checks publicly visible information — the same data any browser or search engine crawler would see. There's nothing private or intrusive about the scan.