Can an AI agent actually do something on your site, such as book an appointment, place an order, click through a configurator or pay? 13 technical checks based on public web agent standards will tell you. Free, and run from Switzerland in line with Swiss data protection law.
Interested in visibility in AI answers instead? Use the AI visibility check (GEO).
Running a shop? Check a single product page with the Shop Check.
Example: www.bombelli-analytics.ch or simply your-company.com. The scan takes about 8 to 15 seconds.
The discoverability checks are visible straight away, each with a code snippet for the fix. Enter your email address to unlock the full report with every category.
Every check plus a code snippet for each recommendation, and the report by email. Enter your email address to unlock it.
The check tests 13 public standards that show whether an AI agent can work with your site, plus three commerce standards. Every check has its own RFC or specification as a source, linked in the detailed finding on each card.
The capabilities layer is the part that goes beyond reading. MCP Server Card, Agent Skills, API Catalog (RFC 9727), OAuth Discovery and WebMCP describe which functions an agent may call on your site. This is exactly where the Agent Readiness Check differs from the AI visibility check, which measures visibility in AI answers.
This is a bet on the future. These standards are young and rarely used. Among the 200,000 most visited domains, Cloudflare counts fewer than 15 with an MCP or API catalogue, and llms.txt has not been ratified yet. If your site passes the check today, you are prepared before agent access becomes the norm.
We test with our own audit logic and give plain explanations and a code snippet for every recommendation. The tool is hosted in Switzerland with cyon in Basel, and data processing follows Swiss data protection law (revDSG). You see the score straight away, and the full report with all snippets once you unlock it.
The score measures how well a website suits AI agents such as ChatGPT, Claude, Cursor or Devin. It covers four categories: Discoverability, Content, Bot Access Control and Capabilities, plus an optional commerce layer.
The AI visibility check measures how visible you are in AI answers from ChatGPT, Claude or Perplexity. The Agent Readiness Check measures whether an AI agent can act on the site, for example book an appointment, place an order, use a configurator or pay. If visibility is what you care about, use the AI visibility check (GEO).
Hosting in Switzerland with cyon in Basel, data processing in line with Swiss data protection law (revDSG), plain explanations and a code snippet for every check. It is one of the few checks that covers the capabilities layer, so it tests whether an agent can act on the site, not only whether it can read it. You see the score and three core checks straight away, and the full report once you enter your email address.
The scanned URL and the score output sit in the server logs only for the duration of the request. There is no tracking, no account and no sharing with third parties. We only collect your email address to unlock the full report.
Yes, completely. The score and the three core checks are visible straight away, and the full report with all code snippets follows once you unlock it. Bombelli Analytics earns its money from the advisory work that may follow, not from the tool itself.
Bombelli Analytics helps Swiss B2B companies make their website usable for AI agents, from discoverability through the capabilities layer to agentic payment. In the report, every measure is backed check by check with the relevant standard and a code snippet.
15 minutes, no obligation. A specialised team from Lucerne for agent-ready websites.