Skip to main content
IAescolaagent-ready · Free tool

What a bot reads when it visits your website

ChatGPT, Claude and Perplexity don't look at your site the way a person does: they ask for specific files and read the reply. Type your domain and we'll make the same requests they do.

Public domains only. Takes about 10 seconds.

The requests we'll make

  1. GEThttps://example.com/

    the homepage, as served

  2. GEThttps://example.com/

    Accept: text/markdown

    the homepage, asked for in Markdown

  3. GEThttps://example.com/robots.txt

    the rules for robots

  4. GEThttps://example.com/sitemap.xml

    the page map

  5. GEThttps://example.com/.well-known/api-catalog

    Accept: application/linkset+json, application/json

    the API catalogue

  6. GEThttps://example.com/.well-known/mcp/server-card.json

    If missing: /.well-known/mcp/server-cards.json · /.well-known/mcp.json

    the MCP server card

  7. GEThttps://example.com/.well-known/agent-skills/index.json

    If missing: /.well-known/skills/index.json

    the index of agent skills

  8. GEThttps://example.com/.well-known/openid-configuration

    If missing: /.well-known/oauth-authorization-server

    how an agent signs in

8 read-only requests to public addresses, plus up to 4 alternatives when one is missing. We change nothing.

Prefer to ask directly? Chat with us on WhatsApp

What it looks at, in four parts

Each request answers a simple question. You don't need to code to understand the result; fixing it sometimes takes a developer.

Discoverability
Can a bot find every page and know where to start?
GET /GET /sitemap.xml
Content
If an agent asks for clean text in Markdown, does it get it, or does it have to decode the HTML?
GET /Accept: text/markdown
Access control
Have you told AI bots what they may read, and what they may use it for?
GET /robots.txt
API, MCP and skills
Can an agent do things on your site (look up, book) and not just read it? Only needed if you offer those services.
GET /.well-known/api-catalogAccept: application/linkset+json, application/jsonGET /.well-known/mcp/server-card.jsonGET /.well-known/agent-skills/index.jsonGET /.well-known/openid-configuration

Why it matters to a small business

People ask before they search

More and more customers ask ChatGPT, Claude or Perplexity which company suits them. These assistants read websites to answer, and what they can't read well they can't cite.

The standards are new

Six of the ten checks look at standards that appeared from 2023 onwards, and some are still drafts. Knowing where you stand now avoids surprises when they settle.

Often it's little work

Many fixes are one text file or a few lines in robots.txt. The advanced ones (MCP, OAuth) only make sense if your site offers services to agents.

Based on public standards: robots.txt (RFC 9309), sitemaps.org, Web Linking (RFC 8288), Content Signals (IETF draft), API Catalog (RFC 9727), MCP, Agent Skills and OAuth/OIDC. Inspired by isitagentready.com.