Built for your kind of site
AIScan fetches your site the way ChatGPT, Claude, Perplexity and shopping agents do, then grades what they find across five dimensions. It reads robots.txt for explicit AI bot rules and Content Signals, follows your XML sitemap, asks for Markdown and /llms.txt, checks for a single H1 and JSON-LD in the server-rendered HTML, and looks for agent endpoints such as an API catalog, an MCP server card and, for stores, UCP and machine-readable pricing.
The pages below turn those checks into a plan for one kind of site. Playbooks are for whoever runs the stack: the WordPress admin, the Shopify store owner, the Next.js or Lovable developer. Each one says which checks that platform usually fails and where the fix lives. Stories are for the business owner: a clinic, a plumber, a restaurant, a studio. Each one names the question a customer asks an AI assistant and the checks that decide whether your site can answer it.
Every page ends with a free scan of your own URL, so you see your score and the failing checks before you change anything. No account needed.
Start from your platform
How your site is built decides where the fix lives. Each playbook shows live averages and the checks that platform fails most often, computed from our own public scans.
Three plugins fighting over robots.txt, and a theme that hides the answer in a tab.
Great structured data out of the box, and a locked-down root that hides the rest.
Server components render beautifully, and the file an agent asks for first does not exist.
The best-rendered HTML we see, and almost nothing at the capability layer.
Shipped in an afternoon, indexed like a single-page app.
A beautiful page built in a canvas, where the root of the domain is not yours to edit.
No detector fired, which means no platform defaults are helping you either.
Start from your business
What an AI agent needs differs by business. Each story names a real moment, the way it currently fails, and the dimensions that decide it. Showing 6 of 40.
Professional services
6Firms sold on expertise, where the agent is summarising your authority.
An assistant is summarising your expertise. It can only summarise what it can read.
Also covers: law firm, solicitors, attorney, barrister chambers, legal practice, conveyancing
Deadline season, a specific question, and a fee page that says 'bespoke'.
Also covers: accountant, bookkeeper, cpa, chartered accountant, tax adviser, payroll bureau
A gorgeous site that says nothing a machine can repeat back.
Also covers: marketing agency, brand studio, design studio, creative agency, web design agency, seo agency
A buyer with six listing tabs open, asking an assistant to do the comparing.
Also covers: estate agent, letting agent, realtor, property management, rental agency, commercial property
A comparison question where every competing answer is a structured product.
Also covers: insurance broker, commercial insurance, liability cover, business insurance, insurance agency, mortgage broker
Job boards are the most machine-shaped content there is — and the most often hidden.
Also covers: recruiter, recruitment agency, staffing agency, headhunter, executive search, temp agency
Scan your site and see where you actually stand
Free, no account, about twenty seconds. The report names every failing check by ID, shows the evidence we found, and gives the fix for your platform — plus a hand-off prompt you can paste straight into Claude Code or Cursor.
npx aiscan-cli yoursite.comPrefer to read first? Browse every check we run or the guide library.