ChatGPT, Claude and Gemini now search, compare and buy for their users. Get a free score for how well your site can be read and used by AI agents, with the fixes that matter most.
Free. Reads only public pages and files. Nothing on your site is changed.
Any public website. No sign-up, no install, and nothing is added to your site.
About 20 read-only requests to the files AI agents look for: robots.txt, llms.txt, structured data, API and MCP discovery.
A score out of 100 in seconds with the biggest gaps first, then the full report and a branded PDF by email.
robots.txt, AI assistant and crawler rules, Content-Signal preferences, XML sitemap.
llms.txt and its format, llms-full.txt, Markdown versions of pages, content readable without JavaScript.
Valid schema.org JSON-LD, Organization and WebSite, products, services and FAQs.
OpenAPI and API catalogue, MCP server discovery, OAuth discovery, NLWeb, agent skills and WebMCP.
Shopify and WooCommerce product feeds and product markup. Only scored for online stores.
Assistants can read your site and find a way to act on it. Keep it current as the standards move.
Assistants can read you but can't do much. Start with the agent-capabilities checks: they carry the most points.
Assistants struggle to read the site at all. Crawl access and llms.txt are the quickest wins.
The Agent-Ready Audit turns your report into a plan, then we build it so assistants like ChatGPT and Claude can search, book and buy on your site.
We turn your report into a build plan, ordered by what moves the score and your business most.
llms.txt, structured data, Markdown pages and an MCP server with sign-in, rate limits and logging.
We host and maintain it, and keep it current as the standards change.
Every check with its result and the exact fix, the fixes ranked by the points they would add, and a branded PDF you can forward to your developer or agency.
The report is yours to keep and act on. If you'd like help, reply to the report email and we'll scope an Agent-Ready Audit with you.
AI assistants such as ChatGPT, Claude and Gemini now browse, compare and buy for their users. An agent-ready site lets them read your content reliably and, ideally, act on it: search your catalogue, check availability, book or order through an API or an MCP server.
The Model Context Protocol is the open standard AI assistants use to call tools. A business that publishes an MCP server lets assistants take real actions on its site, with sign-in, rate limits and logging, instead of guessing from web pages.
No. It only reads public pages and files that any browser or search crawler can read, such as your homepage, robots.txt and llms.txt. It makes about 20 requests and identifies itself as VeluraAgentReadyBot.
There's no single standard yet. Our score weights what matters most for agents acting on a business (APIs and MCP) more heavily than pure content checks, and doesn't penalise non-shops for missing product feeds.