Agentic Readiness

Markdown Rendering

What it checks

Tests whether your site can return content as plain Markdown when an AI agent requests it via the Accept: text/markdown HTTP header.

Why it matters

AI agents and LLMs process Markdown far more efficiently than raw HTML. When your server responds with clean Markdown, agents can understand your content without parsing through navigation menus, ads, scripts, and layout markup. This leads to more accurate citations and better representation in AI-generated answers.

How to improve

• Configure your web server or CMS to detect the Accept: text/markdown header and return a Markdown version of the page content.
• If you use a static site generator (Next.js, Astro, Hugo), add a middleware or route handler that converts your content to Markdown on-the-fly.
• Ensure the response sets Content-Type: text/markdown so agents can confirm the format.
• At minimum, strip navigation, footer, sidebar, and script elements — return only the article body as Markdown.

Scoring

llms.txt

What it checks

Checks whether your site publishes /llms.txt and /llms-full.txt files at the domain root — a proposed standard for communicating site context to AI agents.

Why it matters

Just as robots.txt tells search crawlers how to interact with your site, llms.txt tells AI agents what your site is about, what content is available, and how to navigate it. This helps agents provide more accurate, up-to-date information about your brand and products.

How to improve

• Create a /llms.txt file at your domain root with a brief summary of your site, its purpose, key topics, and links to important sections.
• Create a /llms-full.txt with expanded content — full descriptions, product details, FAQ answers, and any information you want AI agents to reference.
• Both files should be plain text (Content-Type: text/plain).
• Update these files whenever your site's core content or structure changes significantly.

llmstxt.org specification →

Scoring

Structured Data

What it checks

Scans your page for JSON-LD and microdata markup using schema.org vocabularies — the structured metadata that helps machines understand your content type and properties.

Why it matters

Structured data gives AI agents explicit signals about what your content is (an article, FAQ, how-to guide, product review) and its properties (author, date, topics). Without it, agents must infer content type from context, which is less reliable. Rich structured data also improves search engine rich results.

How to improve

• Add JSON-LD script blocks to your pages with relevant schema.org types: Article, BlogPosting, FAQPage, HowTo, BreadcrumbList, etc.
• Include key properties: headline, author, datePublished, dateModified, description, and image.
• For FAQ pages, use FAQPage schema with Question and Answer items.
• Use Google's Rich Results Test to validate your markup.
• Most CMS platforms (WordPress, Ghost, Webflow) have plugins or built-in support for structured data.

schema.org documentation →

Scoring

Content Accessibility

What it checks

Evaluates semantic HTML quality: heading hierarchy (h1 → h2 → h3 without gaps), image alt text coverage, meta description, and Open Graph tags.

Why it matters

Semantic HTML is the foundation of machine-readable content. A clean heading hierarchy lets agents understand your content outline. Alt text on images provides context that agents can reference. Meta descriptions and OG tags give agents concise summaries to cite. These same signals also improve SEO and accessibility for screen readers.

How to improve

• Ensure headings follow a logical hierarchy — don't skip from h1 to h3. Each page should have exactly one h1.
• Add descriptive alt text to every image. Describe what the image shows, not just 'image' or 'screenshot'.
• Include a meta description tag that concisely summarizes the page content (120-160 characters).
• Add Open Graph tags (og:title, og:description, og:image) so agents and social platforms can generate accurate previews.

Scoring