Can AI systems use your website?
Crawlers discover and retrieve content. Browser agents interact with rendered pages for users. We measure the signals each access mode can use.
The shift
IsAgentReady measures machine-readable web signals used across discovery, structured data, semantics, protocols, and security. These static signals do not predict citation or agent task outcomes.
GPTBot, ClaudeBot, Amazonbot, Google-Extended, and 9 more
Across 5 weighted categories
What we measure
Four purpose profiles interpret measured stable-core coverage across five categories. Weights and grade bands are explicit versioned policy choices.
AI CONTENT DISCOVERY
Discovery has the highest configured weight in this methodology. We measure robots.txt directives for 13 named AI user agents, sitemap presence, and llms.txt as a publisher-provided entry point.
- robots.txt
- AI Crawler Directives
- XML Sitemap
- llms.txt
- Meta Robots
- HTTP Bot Access
- Content Freshness
AI SEARCH SIGNALS
Structured data makes your content unambiguously machine-readable. JSON-LD with Schema.org vocabulary tells agents and search engines exactly what your page represents: product, article, organization, or service. Its benefit for AI search is indirect, via the search indexes that feed products like ChatGPT and Perplexity. We validate 17 high-value schema types and check for proper entity linking.
- JSON-LD
- Schema.org Types
- Entity Linking
- Schema Validation
- Organization Schema
- FAQPage Schema
- Author Attribution
CONTENT & SEMANTICS
Browser agents may combine the rendered DOM, accessibility tree, and screenshots. Fetch-only crawlers and non-rendered retrieval depend on accessible HTML. We validate heading hierarchy, semantic markup, and meaningful server-rendered content.
- SSR Detection
- Heading Hierarchy
- Semantic HTML
- ARIA Landmarks
- Alt Text
- Language
- Link Text
- Question Headings
AGENT PROTOCOLS
We measure supported agent interface and discovery signals such as A2A Agent Cards, MCP Discovery, OpenAPI, and agents.json. Their weights are versioned policy choices.
- A2A Agent Card
- MCP Discovery
- OpenAPI
- agents.json
SECURITY & TRUST
We measure HTTPS and security-header signals such as HSTS, CSP, and Referrer-Policy. The category summarizes observed coverage and does not predict whether an agent will trust or use the site.
- HTTPS
- HSTS
- CSP
- X-Content-Type-Options
- X-Frame-Options
- CORS
- Referrer-Policy
Experimental WebMCP evidence
WebMCP forms, JavaScript tools, and browser observations are reported separately as experimental evidence. They do not affect the stable score.
We test a fetchable HTML signal
Our scanner fetches raw HTML without executing JavaScript. This measures what fetch-only crawlers and non-rendered retrieval can access, while browser agents may also use rendered page signals.
Fetchable HTML perspective
Our scan evaluates the initial HTML response. It does not simulate every agent: browser agents may combine the rendered DOM, accessibility tree, and screenshots, while fetch-only crawlers and non-rendered retrieval may not execute JavaScript. Serve important content in fetchable HTML, then test interactive journeys in a browser.
Every AI bot, individually
We parse your robots.txt for all 13 major AI user agents, both crawlers (GPTBot, Google-Extended, Bytespider) and search/assistant bots (ChatGPT-User, ClaudeBot, PerplexityBot). Each has different crawling behavior and purpose.
Emerging protocols
We report supported A2A Agent Card, MCP discovery, and agents.json signals. WebMCP observations are shown separately as experimental, non-scored evidence.
Actionable code snippets
Every failed checkpoint comes with a prioritized recommendation and copy-paste code to fix it. Your report includes per-checkpoint pass/fail results, specific fixes, and a radar chart of your five category scores.
Scan and fix from your editor. Install the MCP server to scan, and AI skills to fix.
claude mcp add isagentready-mcp -- npx -y isagentready-mcp
/plugin marketplace add bartwaardenburg/isagentready-skills
How we score
Multiple checkpoints across 5 categories earn points toward the overall 0-100 score. Weights and grade bands are explicit versioned policy choices.
Methodology 2026-09-02
The legacy overall score remains the weighted five-category score. Four purpose-specific profiles use checkpoint-level earned and available points: AI search visibility, browser agent usability, API and tool readiness, and agentic commerce. Standard, supported, and emerging evidence contributes by default. Experimental checks remain visible but do not affect stable profile scores. Skipped checks stay outside denominators, and commerce is marked not applicable when no commerce check applies.
Profile checkpoint mapping
Current default profile evidence is listed explicitly. Experimental checks remain visible as informational evidence and do not affect the score.
- AI search visibility
- 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 1.10, 1.11, 1.12, 1.13, 2.1, 2.2, 2.3, 2.4, 2.6, 2.7, 2.8, 2.9, 3.1, 3.2, 3.3, 3.5, 3.6, 3.7, 3.12
- Browser agent usability
- 1.13, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7, 3.9, 3.10, 3.11, 3.12, 3.13, 3.14, 3.15, 5.1, 5.3, 5.4, 5.5, 5.6, 5.9
- Informational: 4.1, 4.7, 4.9, 5.11. Does not affect the score.
- API and tool readiness
- 1.9, 1.12, 1.13, 1.14, 4.3, 4.4, 4.5, 4.10, 4.11, 4.17, 4.18, 4.19, 4.20, 4.21, 4.22, 4.23, 4.24, 5.1, 5.3, 5.4, 5.5, 5.6, 5.7, 5.9, 5.12, 5.13, 5.14
- Informational: 4.1, 4.6, 4.7, 4.9, 5.10. Does not affect the score.
- Agentic commerce
- 2.3, 2.9, 4.5, 4.17, 4.22, 5.1, 5.3, 5.4, 5.5, 5.6, 5.9, 5.12, 5.13, 5.14
- Informational: 4.12, 4.13, 4.14, 4.15, 4.16, 5.10, 5.11. Does not affect the score.
Letter grades
Category weights
How each category contributes to the overall score
Built by a practitioner
IsAgentReady exists because the standards are moving fast and most websites are being left behind.
Bart Waardenburg
AI Agent Readiness Expert - The Hague, NL
The web is going through one of its biggest shifts since mobile. AI agents are starting to browse on behalf of people, and new standards like MCP, WebMCP, and llms.txt are rewriting how that works. I find it genuinely fascinating to dig into these protocols and figure out what they mean for the websites we build. IsAgentReady started from that curiosity: making this shift concrete and measurable, so businesses can see where they stand and what to improve.
Dive deeper
Learn how ChatGPT, Google, and Claude decide which websites to cite in their AI-generated answers.
Does Schema Markup Get You Cited by AI? What the Data Actually Shows
A 2026 Ahrefs study tracked 1,885 pages adding schema markup, and AI-search citations barely moved. We separate correlation from causation: what structured data is actually proven to do, what really drives citations, and where to spend your effort.
Content Negotiation for AI Agents: Why Sentry Serves Markdown Over HTML
Sentry co-founder David Cramer shows how content negotiation — a 25-year-old HTTP standard — saves AI agents 80% of tokens. We break down the implementation: Accept headers, markdown delivery, authenticated page redirects, and what this means for every website preparing for agent traffic.
Cloudflare /crawl Endpoint: One API Call to Crawl Any Website
Cloudflare launched a /crawl endpoint that crawls entire websites with one API call — returning HTML, Markdown, or AI-extracted JSON. We break down what this means for AI agent readiness: why your robots.txt, sitemap, semantic HTML, and server-side rendering now matter more than ever.
Frequently asked questions
Common questions about AI agent readiness, how we measure it, and what you can do to improve your score.
Explore more
Measure applicable web signals and get a prioritized action plan.