Quickstart

To check whether a site is ready for AI agents, open the AI Site Scanner, type the domain or a page URL into the box, and click Scan site. You get a letter grade, a score out of 100, a score for each category, and the most important failed checks, each with a prompt you can paste into an AI assistant to fix it.

How do I scan a site, step by step?

1. Enter the URL. Type a domain such as example.com or a full page URL such as https://example.com/pricing. Without a scheme, https:// is assumed. The site has to be public: private network addresses and localhost are refused.
2. Pick what to grade. Specific page (the default) grades the URL you entered plus the site-wide files agents look for. Whole site also reads the sitemap and grades up to 5 of the site's pages. Page scan vs whole-site scan explains the difference.
3. Pick who can see the report. Public (the default) gives the report a shareable link and lists it in Recently scanned. Private keeps it out of that list and is a Toolz Pro feature. See Public and private scans, limits and Pro.
4. Click Scan site. The button reads "Scanning…" while Toolz.dev's server fetches the site. The page itself gets up to 15 seconds to respond, and most discovery files are requested in parallel with a 2.5 second limit each, so a slow site makes for a slow scan.

Optional: open Customize scan before step 4 to limit the scan to content-site checks, API checks or a hand-picked set. Scores, grades and scan filters covers what that does to the score.

How do I read the report?

The report has up to four parts, top to bottom:

Part	What it shows
Score header	The URL scanned, the grade (A+ to F), the score out of 100, the detected platform (WordPress, Shopify, Next.js and others), the time to first byte, and the number of critical issues
Category cards	One card per category that ran, with its own 0 to 100 score and a "passing · to fix" count
Pages graded	Only on a whole-site scan: each page from the sitemap with its score, grade and failing checks, worst first
Prioritized issues	Up to six failed checks, critical first, each with what was found and a Copy fix prompt button

Critical issues are failed checks with a weight of 7 or 8, warnings have a weight of 4 to 6, and the rest are info. When nothing is failing, the report says "No blocking issues found. This site is in good shape for AI agents."

Under the report, Open this report on its own page takes you to the stored copy at /ai-scan/r/<id>, which is the link to keep or send. Reports, sharing and history covers it.

How many scans can I run?

The line under the controls shows it, in the form "N of M free scans left today", and when you are not signed in it also says how many a day a free account gets. The numbers are read from the usage service, not written into the page, so they match what the server will allow. With Toolz Pro the line reads "Unlimited scans with Pro". When the allowance is used up, the scanner says "You have reached your daily scan limit. Sign in or upgrade for more."

What next?

• Fix the issues: Fixing issues with fix prompts shows how to use the prompts and confirm the change with a re-scan.
• Wondering why a check failed? What the scan checks gives the exact pass condition for each.
• Want the same scan in a terminal or CI? The command line runs it with npx @toolz-dev/aiscan example.com.