# AgentsReadable — Agent-Ready Business Infrastructure We make businesses discoverable, understandable, trusted and actionable for AI agents: one canonical record of business facts, compiled into the website, structured data, an agent API and protocol adapters — with honest measurement of what AI says about you. ## Readiness levels - **Level 0 · Invisible** — Agents can't reliably fetch or read your site. - **Level 1 · Discoverable** — Agents can reach and index your pages. robots.txt and sitemap let them in. - **Level 2 · Understandable** — Structured data says who you are, what you offer and at what price. - **Level 3 · Trusted** — Your facts agree everywhere — site, data, profiles, AI answers — and independent sources back them up. - **Level 4 · Recommendable** — Agents have the decision data to shortlist you: who it's for, when it fits, how it compares. - **Level 5 · Actionable** — Agents can query live facts and send requests: quotes, availability, sign-ups. - **Level 6 · Transactable** — Agents can complete a booking, order or subscription through a documented path. - **Level 7 · Agent-native** — Agents can use what you do directly, as a tool — and you can measure them as a channel. The instant check measures levels 0–2. Higher levels need the full audit. ## What the check covers 31 checks across 7 layers. - **Discovery** (10): HTTPS · Content without JavaScript · robots.txt · AI crawler access · Google-Extended · Sitemap · No noindex · Canonical URL · Title and description · CDN/firewall blocking signal - **Semantics** (4): Valid JSON-LD · Organization · Offering type (Service, Product, App, Hotel…) · WebSite - **Entity** (5): Consistent name · Machine-readable contact · Official profiles · Logo · Location or service area - **Facts** (4): Prices in data · Prices on the page · Data vs page match · Policy links - **Freshness** (1): Last-modified dates - **Agent interfaces** (4): Markdown version · API catalog / OpenAPI · llms.txt · Agent protocol endpoint (MCP, UCP, A2A) - **Actionability** (3): Detectable main action · Labeled form fields · No CAPTCHA on step one Only in the full audit: Trust · Reputation · Transactions. ## Offer and pricing - **Agent Readiness Audit** — $490. A complete audit of how agents find, read and act on your business, with a fix plan you can hand to any team. Delivered in 3 business days. - All 10 layers, including Trust, Reputation and Transactions - Every instant-check finding verified by hand - AI answers sampled across several engines, over several days - Fact-by-fact comparison: site, structured data, profiles, AI answers - Prioritized fix plan with effort levels - Walkthrough call with the engineer who ran it - **Fix sprint** — from $990. We implement the audit's fix plan on your current site: structured data, crawler access, price and entity consistency, Markdown and llms.txt. - **Agent-Ready Website** — from $2,900. A new or rebuilt site compiled from one canonical fact record, synced from the systems you already use — when the audit shows a rebuild is the cheaper path. - **Agent API and adapters** — from $4,900. A documented API for live facts and actions — quotes, availability, booking or sign-up requests — plus MCP, OpenAPI and feed adapters where available. - **Monitoring** — from $149/mo. Daily re-checks, AI answer sampling, fact-drift alerts, new agent channels as they open, and the agent dashboard. ## FAQ ### Is this SEO? Partly overlapping, with a different goal. SEO optimizes ranked links that people click. We make your business facts correct, verifiable and actionable for software agents. Good structured data helps both, but we don't sell rankings. ### What exactly does the free check do? It fetches your homepage and key public pages, reads robots.txt and your sitemap, parses structured data, asks one AI engine what it knows about you, and compares the facts. That's 31 checks across 7 layers. It never logs in, submits forms or changes anything. ### How is the score calculated? Each of the 7 layers scores 0–100 from its checks, and the overall score combines them with published weights. The AI snapshot is reported separately and never changes the score. The full method is published at /method. ### Why does the AI snapshot differ from what I see? AI answers vary by engine, time, location and conversation. The snapshot is one question at one moment, shown with its timestamp. The full audit samples several engines over several days. ### What do I get for $490? A manual audit of all 10 layers, AI answers sampled across several engines, a fact-by-fact comparison and a prioritized fix plan with effort levels, plus a walkthrough call. Fixed price, delivered in 3 business days. You're free to implement the plan yourself. ### Will AI agents be able to book or buy from us? Where platforms allow it. Agentic commerce is rolling out by region and category, retail first. We build documented action endpoints on a protocol-independent core, so you can connect as channels open — without betting on a single platform. ### Do you store the address I check? We store the address, the result and the AI snapshot so you can come back to them. Results are kept for 12 months. We only check public pages, and we never store page content beyond the evidence lines you see. ### Do I need developers? Not for the check or the audit. For builds, we work with your team or on our own. Most outputs need access to your hosting, DNS or CMS rather than new code. ### What are protocol adapters? Small, standard interfaces that let agents query your facts or take actions: an MCP server, an OpenAPI description, llms.txt, and product or availability feeds. They all read from the same canonical record, so they never disagree with each other. ## For agents - Canonical fact record (business, offerings, prices, FAQ): /facts.json - This page as Markdown: /index.md - API description (OpenAPI): /openapi.json - Guide for language models: /llms.txt - Scoring method: /method - Run an agent readiness check: POST /api/checks - Request an audit (a human confirms by email): POST /api/audit-requests ## Contact - Email: hello@agentsreadable.com - Profile: https://github.com/agentsreadable - Privacy: /privacy - Terms: /terms Updated 2026-10-06 · SH & SH Co., Ltd.