Is your site agent-ready?
ChatGPT, Claude, and Perplexity are the new front door. AI Site Scanner grades how well your site talks to them across 31 signals - and hands you the exact fixes.
Grades the URL you enter, plus the files agents look for.
Unlike most Toolz tools, the AI Site Scanner runs on our server - fetching another site is not something a browser can do. Only the URL you enter is sent.
Three steps to your grade
Drop in a URL
Any public website. We fetch the homepage plus every discovery file an agent would look for.
We probe 31 signals
robots.txt, llms.txt, Markdown negotiation, .well-known endpoints, OAuth metadata, agent cards, and more.
Ship the fixes
Review the report, copy the fix prompts into your AI assistant, deploy, and re-scan to confirm.
Recently scanned
View all public scans →| Site | Grade | Score | Platform | Scanned |
|---|---|---|---|---|
| caep.org.ly | F | 28/100 | Unknown / custom | Nov 9 |
| languageclub.es | D | 64/100 | WordPress | Nov 8 |
| avito.ma | F | 9/100 | Unknown / custom | Nov 1 |
| portal.quanta5.co.za | F | 21/100 | Unknown / custom | Oct 29 |
| miramarltd.com | F | 41/100 | WordPress | Oct 26 |
| pxld.sh | F | 49/100 | Unknown / custom | Oct 21 |
| wpstats.dev | D | 64/100 | Unknown / custom | Sep 27 |
| jekanov.com | C | 70/100 | Unknown / custom | Sep 26 |
| master-addons.com | C | 74/100 | WordPress | Sep 26 |
| wpadminify.com | C | 70/100 | Next.js | Sep 26 |
The newest public scan of each site. Scans run with the switch set to private never appear here.
Built for the agent web
Free, no signup
Scan any public site in seconds. No credit card, no email gate - just a URL and a grade.
Cross-agent detection
Evaluates what GPTBot, ClaudeBot, PerplexityBot, Google-Extended, and CCBot actually encounter at your origin.
Platform-aware
Fingerprints WordPress, Shopify, Next.js, Webflow, Framer, Wix, Bolt, and Lovable - fixes speak your stack's language.
A report you can share
Each scan saves its report: the URL, the grade and every check result, at a link you can share. The page's HTML is not kept.
Always current
Tracks emerging standards as they ship: llms.txt, Content Signals, Web Bot Auth, DNS-AID, MCP, A2A, x402.
Plain-English fixes
Every failing check comes with a ready-to-paste prompt for your AI coding assistant. No spec-reading required.
Use AI Site Scanner from your own agent
The same report is available from a terminal, from your own code and soon from your assistant. Every route runs the same checks and returns the same score.
Command line
Scan a site from a terminal or a CI job. Published on npm, MIT-licensed, no dependencies, and it fails a build below the score you set.
npx @toolz-dev/aiscan example.com # fail a build under a score npx @toolz-dev/aiscan example.com --min-score 80
HTTP endpoint
The full report as JSON, from one request. No key and no signup, so your own tooling can grade a site in a few lines.
GET https://toolz.dev/api/ai-scan?url=example.com
# or
POST https://toolz.dev/api/ai-scan
{ "url": "example.com" }MCP serverNot released yet
A scan_site tool for Claude, Codex, Gemini CLI and Cursor, so an assistant can grade a site and fix what it finds in your repo.
Read the docsBuilt for everyone shipping the agent web
Site owners
Find out in one scan whether the AI era can see your website at all.
Developers
A standards checklist with copy-paste fixes and a JSON API.
Indie hackers & SaaS founders
Get your product cited when users ask ChatGPT or Claude for recommendations.
SEO & AEO consultants
Add agent-readiness audits to your offering with shareable reports.
Agencies
Scan client sites, track grades over time, and prove the fixes you ship.
Ecommerce brands
Prepare for agentic commerce: x402, ACP, and structured product data before competitors do.
What each plan changes about a scan
Every plan grades a single page for free. Pro adds the whole site and private reports; Advanced adds the keys that let your own code and your assistant run the same scan.
| What you get | Free$0 | Pro$9/mo | Advanced$24/mo |
|---|---|---|---|
| Scan a single page | |||
| Scans per day | 10 a day | Unlimited | Unlimited |
| Whole-site scan, up to 5 pages | - | ||
| Private reports, kept out of the directory | - | ||
| Score history for a domain | - | ||
| API keys for the HTTP endpoint | - | - | |
| MCP server for Claude, Codex and Cursor | - | - |
Prices are billed monthly and drop on a yearly plan. See full pricing
FAQ
Find out what the agents see
One URL, a few seconds, and a prioritized list of what to fix.
Scan a site