# AgentReady Monitor > What do AI models recommend when buyers ask about your category? AgentReady Monitor measures every month what ChatGPT, Claude, Perplexity and Gemini answer when potential buyers ask about your category. Score, share of answer, competitor tracking and alerts — for companies that want to know whether they show up in AI answers at all. ## The problem ### Google rankings no longer tell you whether you get recommended Buyers do not open a results page any more, they ask a model: "Which CRM fits a trade business with 20 people?" What comes back as a shortlist decides your pipeline — and nobody measures it. That is where the Monitor starts. Every month it asks the same 25 questions your buyers would ask, across every relevant model, and records who gets recommended, what is claimed about you, and where a gap opens up. ## How it works ### 01 Create a case Domain, category, ideal customer, 3–5 competitors, 2–3 use cases. Four minutes. Plus a fact sheet with your real prices — without that ground truth there is no way to judge whether a model tells the truth about you. ### 02 Measure the baseline We generate 25 buying prompts in five blocks: category shortlist, facts about you, use case, buying criteria, head-to-head against your competitors. Every prompt runs against every model in every language — with live web search, the way your buyer sees it. ### 03 Have it scored Every answer is judged against your fact sheet: visibility (0–2), accuracy (0–2), sentiment (−1 to +1), the dominating competitor, and the open gaps. That yields a score from 0 to 100, per model and overall. ### 04 Repeat every month The run repeats automatically. You see the trend, the delta to last month, and get a report by email. If the score drops by more than 10 points or a new competitor shows up in the recommendations, you get an alert. ## What you get ### Score 0–100 One value per model and one overall, built from visibility, accuracy and sentiment. Trend chart across every month measured. ### Share of answer The share of comparison questions where you are named — next to every competitor. "Craftnote is recommended in 80% of comparisons, you in 40%." ### Answer matrix Every single answer verbatim, sorted by prompt and model, with its score and reasoning. Filterable by block and language. ### Gaps What the models could not answer about you — deduplicated and ranked. That is your FAQ backlog for the next few weeks. ### Alerts Email when the score drops by more than 10 points or a competitor newly enters the recommendations. Pro plan. ### API instead of click paths Everything the dashboard does, the API does too. Bearer tokens, an OpenAPI spec, one endpoint per job — built for agents that ask on their own. ## Models measured Measured through OpenRouter with live web search — the mode your buyer actually uses in a chat window. Four core models in every paid plan, all nine on Pro, including the assistants that dominate in Asia. - OpenAI GPT-6 Astra - Anthropic Claude Sonnet 5 - Google Gemini 3.8 Flash - Perplexity Sonar Pro - xAI Grok 4.6 - DeepSeek V4 Pro - Qwen3.8 Max - MoonshotAI Kimi K3 - Z.ai GLM 5.3 ## Pricing ### Trial — 0 EUR (14 days) - 14 days free, no credit card - 1 case, 25 prompts, baseline run - 2 models (ChatGPT, Claude), English - Score, share of answer, gaps - Cases: 1 - Models: 2 - Languages: en - Alerts: no - Share links: no - Manual re-runs: no ### Starter — 49 EUR/month - 1 case, monthly measurement run - 4 core models: ChatGPT, Claude, Gemini, Perplexity - Sources: which sites the models cite about your category - English + German - Monthly report by email - Cases: 1 - Models: 4 - Languages: en, de - Alerts: no - Share links: no - Manual re-runs: no ### Pro — 149 EUR/month - 3 cases (e.g. several products or markets) - All 9 models: the 4 core ones plus Grok, DeepSeek, Qwen, Kimi and GLM - The Asian assistants your buyers there actually use - Alerts on a score drop over 10 or a new competitor - Public share links for reports - Manual re-runs, any time - Cases: 3 - Models: 9 - Languages: en, de - Alerts: yes - Share links: yes - Manual re-runs: yes All prices net per month, cancel any time. No setup fee, no minimum term. ## FAQ ### How is this different from SEO tools? SEO tools measure rankings in a list of results. We measure answers. There is no position 1 to 10, there is prose in which three vendors appear and the rest does not exist. So we measure mention, accuracy and sentiment — not position. ### Why do I need a fact sheet? Without ground truth there is no way to judge whether an answer is right. Once your prices, plans and audience are on file, every answer can be checked against them — and you see immediately when a model quotes an outdated price. That is why prices and plans are mandatory. ### How often do you measure? Automatically once a month, plus a baseline run right after you create the case. On the Pro plan you can measure again at any time — after a relaunch, for example. ### Which models do you query? On the trial, ChatGPT and Claude. From Starter, Perplexity and Gemini as well, plus a second language. All with live web search, so the answer matches what a buyer sees in a chat. ### How are the answers scored? A separate model without web search judges every answer strictly against your fact sheet and competitor list. It gets four numbers to fill in: visibility 0–2, accuracy 0–2, sentiment −1 to +1, and the dominating competitor. Score = visibility × 25 + accuracy × 15 + (sentiment + 1) × 5, normalised to 0–100. ### How does this relate to the AgentReady Check? The Check measures the technology of a domain: robots.txt for AI crawlers, llms.txt, structured data, server-rendered HTML. The Monitor measures the outcome: what the models actually answer. Every monthly run triggers a fresh Check scan, so both numbers sit side by side — and a drop in visibility gets a cause instead of a shrug. ### Is there an API? Yes. Every dashboard function exists as an endpoint under /api/v1, authenticated with a bearer token. The OpenAPI spec lives at /api/v1/openapi.json, the documentation at /docs/api. An MCP server mirroring exactly these endpoints runs at /api/mcp — v0, running inside this application; it is planned to move to a standalone Cloudflare Worker without changing the URL or the tool names. ### Can I cancel? Any time in the customer portal, effective at the end of the current period. The trial runs for 14 days and needs no credit card. ## Links - Web: https://ai-agent-ready.com/ - Pricing: https://ai-agent-ready.com/pricing - FAQ: https://ai-agent-ready.com/faq - Fix: https://ai-agent-ready.com/fix - Blog: https://ai-agent-ready.com/blog - API docs: https://ai-agent-ready.com/docs/api - OpenAPI: https://ai-agent-ready.com/api/v1/openapi.json - Agent manifest: https://ai-agent-ready.com/.well-known/ai-plugin.json - Technical check: https://ai-agent-ready.com/check ## More from The Autopilot - [The 21-Process Check](https://check.the-autopilot.com): How much of your business already runs without you? (German) - [Clapr](https://www.clapr.io): Run your YouTube pipeline on autopilot with AI agents. - [Loop Engineer](https://www.loop-engineer.de): Learn loop engineering in eight missions. (German) A product by The Autopilot — https://the-autopilot.com --- # Pricing — AgentReady Monitor > What do AI models recommend when buyers ask about your category? ### Trial — 0 EUR (14 days) - 14 days free, no credit card - 1 case, 25 prompts, baseline run - 2 models (ChatGPT, Claude), English - Score, share of answer, gaps - Cases: 1 - Models: 2 - Languages: en - Alerts: no - Share links: no - Manual re-runs: no ### Starter — 49 EUR/month - 1 case, monthly measurement run - 4 core models: ChatGPT, Claude, Gemini, Perplexity - Sources: which sites the models cite about your category - English + German - Monthly report by email - Cases: 1 - Models: 4 - Languages: en, de - Alerts: no - Share links: no - Manual re-runs: no ### Pro — 149 EUR/month - 3 cases (e.g. several products or markets) - All 9 models: the 4 core ones plus Grok, DeepSeek, Qwen, Kimi and GLM - The Asian assistants your buyers there actually use - Alerts on a score drop over 10 or a new competitor - Public share links for reports - Manual re-runs, any time - Cases: 3 - Models: 9 - Languages: en, de - Alerts: yes - Share links: yes - Manual re-runs: yes All prices net per month, cancel any time. No setup fee, no minimum term. Start here: https://ai-agent-ready.com/app/start A product by The Autopilot — https://the-autopilot.com --- # FAQ — AgentReady Monitor ### How is this different from SEO tools? SEO tools measure rankings in a list of results. We measure answers. There is no position 1 to 10, there is prose in which three vendors appear and the rest does not exist. So we measure mention, accuracy and sentiment — not position. ### Why do I need a fact sheet? Without ground truth there is no way to judge whether an answer is right. Once your prices, plans and audience are on file, every answer can be checked against them — and you see immediately when a model quotes an outdated price. That is why prices and plans are mandatory. ### How often do you measure? Automatically once a month, plus a baseline run right after you create the case. On the Pro plan you can measure again at any time — after a relaunch, for example. ### Which models do you query? On the trial, ChatGPT and Claude. From Starter, Perplexity and Gemini as well, plus a second language. All with live web search, so the answer matches what a buyer sees in a chat. ### How are the answers scored? A separate model without web search judges every answer strictly against your fact sheet and competitor list. It gets four numbers to fill in: visibility 0–2, accuracy 0–2, sentiment −1 to +1, and the dominating competitor. Score = visibility × 25 + accuracy × 15 + (sentiment + 1) × 5, normalised to 0–100. ### How does this relate to the AgentReady Check? The Check measures the technology of a domain: robots.txt for AI crawlers, llms.txt, structured data, server-rendered HTML. The Monitor measures the outcome: what the models actually answer. Every monthly run triggers a fresh Check scan, so both numbers sit side by side — and a drop in visibility gets a cause instead of a shrug. ### Is there an API? Yes. Every dashboard function exists as an endpoint under /api/v1, authenticated with a bearer token. The OpenAPI spec lives at /api/v1/openapi.json, the documentation at /docs/api. An MCP server mirroring exactly these endpoints runs at /api/mcp — v0, running inside this application; it is planned to move to a standalone Cloudflare Worker without changing the URL or the tool names. ### Can I cancel? Any time in the customer portal, effective at the end of the current period. The trial runs for 14 days and needs no credit card. More: https://ai-agent-ready.com/ A product by The Autopilot — https://the-autopilot.com --- # AgentReady Check > The 3-minute check: 9 areas, 100 points, your agent-readiness score. What can AI agents actually read about you? - 9 technical checks against your live domain, across four pillars: Retrievable, Explainable, Actionable, Distributed. - Result in under 60 seconds. No login, no setup, free. - A score from 0 to 100 plus one concrete next step per area — and an honest "we could not verify this" where the answer is not visible from outside. Domains scanned so far: 21. ## The 9 checks ### R — Retrievable (51 points) Can an AI agent fetch and read your pages at all? #### AI crawler access (18 points) We parse your robots.txt and additionally fetch your homepage with the real user agents of GPTBot, ClaudeBot and PerplexityBot. If the status codes differ from a browser request, something is blocking you — usually a bot rule in a CDN or firewall. **Why it matters:** Whoever gets locked out does not appear in AI answers. That is the single most expensive mistake on this list. #### JavaScript dependency (14 points) We measure the visible text on your homepage without JavaScript, and how much of the delivered HTML is executable script code. Under 1,500 characters without JavaScript is critical. **Why it matters:** AI crawlers do not execute JavaScript. Anything that only appears in a browser does not exist for them. #### Agent basics (10 points) sitemap.xml, rel="canonical", a findable route to legal notice and contact, homepage response time under 3 seconds. **Why it matters:** The homework. Without it, the rest only gets you so far. #### Search index verification (9 points) We look for Google Search Console and Bing Webmaster verification in your DNS TXT records and your homepage meta tags, and check whether robots.txt declares a sitemap that actually resolves. **Why it matters:** Verified domains get crawled sooner and deeper — and the Bing index is what Microsoft Copilot answers from. Note the honest limit: verification by file upload or Tag Manager is invisible from outside, so we report "could not verify", never "not registered". ### E — Explainable (37 points) Once it has your pages, can it state facts about you instead of guessing? #### llms.txt (14 points) We check whether /llms.txt really answers with 200, serves text, and is not an HTML soft 404 in disguise. /llms-full.txt earns bonus points. **Why it matters:** Your llms.txt is the curated short version: product, value, pricing, links — written for models rather than for people. #### Structured data (14 points) We read every JSON-LD block in your HTML and look at which @types appear — Organization, Product, Offer, FAQPage — and whether prices are stored as numbers. **Why it matters:** Structured data is the difference between a model interpreting you and a model quoting you. #### Feed & product data (9 points) For every product object in your JSON-LD we check the six fields agents rely on: name, price, currency, availability, identifier (GTIN/SKU/MPN) and image — plus whether return and shipping terms are machine-readable. **Why it matters:** "What does it cost, is it in stock, can I send it back?" are the questions that decide a purchase. An agent will not answer them from prose. ### A — Actionable (7 points) Can it do more than describe you — can it act on your offering? #### Agentic buyability (7 points) We detect your shop platform and look for references to an agent-callable checkout in your HTML, robots.txt and llms.txt. **Why it matters:** Reading about your product is not buying it. This is a heuristic and we label it as one — there is no standard way to discover a checkout endpoint from the outside. ### D — Distributed (5 points) Are your products ready for the catalogues agents shop from? #### Catalog prerequisites (5 points) Whether your product data meets the preconditions for merchant catalogues: identifiers on every product, price and availability together, and a verified domain. **Why it matters:** We check prerequisites, not listings. Whether you are actually in Google Merchant Center or the ChatGPT merchant catalogue cannot be queried from outside, so we do not claim to know. ## Score bands - **0–39 (red):** AI agents get no reliable picture of you. Usually there is either blocking or an empty HTML shell behind this. - **40–69 (yellow):** The fundamentals are in place, the rest is missing. Typical case: crawlers get through, but there is neither an llms.txt nor structured data. - **70–100 (green):** You are agent-ready. From here it is no longer the technology that decides, but what is written about you. ## Price Free. No login, no credit card. 3 of 9 checks are visible immediately, the remaining 6 after you enter an email address. ## FAQ **What exactly does the check measure?** 9 technical areas across four pillars — Retrievable, Explainable, Actionable, Distributed — worth 100 points in total: AI crawler access (18), JavaScript dependency (14), Agent basics (10), Search index verification (9), llms.txt (14), Structured data (14), Feed & product data (9), Agentic buyability (7), Catalog prerequisites (5). Everything is measured live against your domain, nothing is estimated. **How can an area be worth nothing?** Because not every area applies to every site. A blog has no checkout for an agent to use, so scoring it zero for that would be a false negative dressed up as a measurement. Areas that do not apply are removed from the calculation entirely — the score is a percentage of what is actually applicable. And where an area does apply but we genuinely cannot see it from the outside, it counts as half rather than zero. **What does it cost?** Nothing. The check is permanently free, with no login and no credit card. 3 of the 9 areas are visible immediately, the remaining 6 after you enter your email address. **Why do I have to give my email for 6 of the checks?** Because we send you the full report by email — all 9 areas with one concrete fix each, in a form you can forward to your engineers. You also join our newsletter and can unsubscribe with one click. **How long does a scan take?** Usually under 60 seconds. Every individual check has a 10-second timeout; a check that fails is left out of the score rather than counted as zero, and never aborts the scan. **Does the scan change anything on my website?** No. We only read publicly available URLs: your homepage, robots.txt, llms.txt, llms-full.txt and sitemap.xml. We additionally look up your public DNS TXT records to check for search-index verification. We write nothing, submit nothing, and register nothing on your behalf. **What happens to my data?** Every scan is stored anonymously — domain, score and check results — and feeds the aggregated market data at /api/v1/stats. Your email address is stored separately and is never published alongside a specific scan. **Is the result public?** The result page is shareable and indexable by default. On the page itself you can switch it to noindex with one click. **Is there an API?** Yes. Everything the interface can do, the API can do: POST /api/v1/scans starts a scan, GET /api/v1/scans/{id} returns the result, GET /api/v1/stats returns aggregated market data. The OpenAPI spec lives at /api/v1/openapi.json. **How is this different from AgentReady Monitor?** The check measures technology: can agents reach your content at all? The Monitor measures effect: do ChatGPT, Claude and Perplexity actually recommend you, and in what words? ## For agents - API documentation: https://ai-agent-ready.com/docs/api - OpenAPI spec: https://ai-agent-ready.com/api/v1/openapi.json - Agent manifest: https://ai-agent-ready.com/.well-known/ai-plugin.json — points at the same OpenAPI, no auth, free. - An MCP server built on this API is in progress; it consumes /api/v1 exclusively. --- A product by The Autopilot — https://the-autopilot.com --- # How it works The 3-minute check: 9 areas, 100 points, your agent-readiness score. No login, no setup, free. ## In three steps 1. **Enter a domain.** `example.com` is enough — we ignore protocol and path. 2. **Wait.** We fetch your homepage, robots.txt, llms.txt, llms-full.txt and sitemap.xml, look up your public DNS TXT records for search-index verification, and additionally test the homepage with the user agents of GPTBot, ClaudeBot and PerplexityBot. Every request has a timeout; nothing is retried more than once. 3. **Read the result.** A score from 0 to 100 across 9 areas in four pillars, with plain-language findings and one concrete next step each. ## The 9 checks in detail ### R — Retrievable (51 points) Can an AI agent fetch and read your pages at all? #### AI crawler access — 18 points We parse your robots.txt and additionally fetch your homepage with the real user agents of GPTBot, ClaudeBot and PerplexityBot. If the status codes differ from a browser request, something is blocking you — usually a bot rule in a CDN or firewall. **Why it matters:** Whoever gets locked out does not appear in AI answers. That is the single most expensive mistake on this list. #### JavaScript dependency — 14 points We measure the visible text on your homepage without JavaScript, and how much of the delivered HTML is executable script code. Under 1,500 characters without JavaScript is critical. **Why it matters:** AI crawlers do not execute JavaScript. Anything that only appears in a browser does not exist for them. #### Agent basics — 10 points sitemap.xml, rel="canonical", a findable route to legal notice and contact, homepage response time under 3 seconds. **Why it matters:** The homework. Without it, the rest only gets you so far. #### Search index verification — 9 points We look for Google Search Console and Bing Webmaster verification in your DNS TXT records and your homepage meta tags, and check whether robots.txt declares a sitemap that actually resolves. **Why it matters:** Verified domains get crawled sooner and deeper — and the Bing index is what Microsoft Copilot answers from. Note the honest limit: verification by file upload or Tag Manager is invisible from outside, so we report "could not verify", never "not registered". ### E — Explainable (37 points) Once it has your pages, can it state facts about you instead of guessing? #### llms.txt — 14 points We check whether /llms.txt really answers with 200, serves text, and is not an HTML soft 404 in disguise. /llms-full.txt earns bonus points. **Why it matters:** Your llms.txt is the curated short version: product, value, pricing, links — written for models rather than for people. #### Structured data — 14 points We read every JSON-LD block in your HTML and look at which @types appear — Organization, Product, Offer, FAQPage — and whether prices are stored as numbers. **Why it matters:** Structured data is the difference between a model interpreting you and a model quoting you. #### Feed & product data — 9 points For every product object in your JSON-LD we check the six fields agents rely on: name, price, currency, availability, identifier (GTIN/SKU/MPN) and image — plus whether return and shipping terms are machine-readable. **Why it matters:** "What does it cost, is it in stock, can I send it back?" are the questions that decide a purchase. An agent will not answer them from prose. ### A — Actionable (7 points) Can it do more than describe you — can it act on your offering? #### Agentic buyability — 7 points We detect your shop platform and look for references to an agent-callable checkout in your HTML, robots.txt and llms.txt. **Why it matters:** Reading about your product is not buying it. This is a heuristic and we label it as one — there is no standard way to discover a checkout endpoint from the outside. ### D — Distributed (5 points) Are your products ready for the catalogues agents shop from? #### Catalog prerequisites — 5 points Whether your product data meets the preconditions for merchant catalogues: identifiers on every product, price and availability together, and a verified domain. **Why it matters:** We check prerequisites, not listings. Whether you are actually in Google Merchant Center or the ChatGPT merchant catalogue cannot be queried from outside, so we do not claim to know. ## How the score is calculated The 9 areas are worth 100 points in total. The score is not a plain sum of those points, though — it is a percentage of the points that actually apply to your domain: ``` score = round(100 × points earned ÷ points applicable) ``` Two rules decide what "applicable" means, and both exist for the same reason: we would rather tell you nothing than tell you something false about your own site. **An area that does not apply is removed from the calculation.** A blog has no checkout an agent could use and no product catalogue to be listed in. Scoring those areas zero would not be a measurement, it would be a penalty for being a blog. So they leave both sides of the fraction, and the remaining areas are scaled back up to 100. **An area we cannot see counts as half.** Some things are genuinely invisible from the outside. If you verified your domain in Google Search Console by uploading an HTML file, or through Google Tag Manager, or your hosting provider did it for you automatically, there is no DNS record and no meta tag for us to find. We will say "we could not verify this", never "you are not registered" — and score it as half rather than zero, so a false negative cannot cost you the full points and doing nothing cannot earn you them either. This is why two sites can both score 70 while a different number of areas were measured. The result page names the areas that were left out and why. - **0–39 (red):** AI agents get no reliable picture of you. Usually there is either blocking or an empty HTML shell behind this. - **40–69 (yellow):** The fundamentals are in place, the rest is missing. Typical case: crawlers get through, but there is neither an llms.txt nor structured data. - **70–100 (green):** You are agent-ready. From here it is no longer the technology that decides, but what is written about you. ## What we do NOT do - We write nothing on your site and submit no forms. - We register nothing on your behalf — no Search Console, no Bing, no IndexNow ping. - We call no authenticated API and read nothing that is not public. - We execute no JavaScript — exactly like the AI crawlers we simulate. - We do not crawl your site in full. It is at most eleven requests, less than a single browser visit. - We do not claim to know whether you are listed in a product catalogue. That cannot be queried from outside, so we check the prerequisites and say so. ## What we store Domain, score and check results are stored anonymously and feed the aggregated market data at `/api/v1/stats`. We only store an email address if you enter one yourself, and we never publish it. ## After that The technical check is half the truth. Whether AI actually RECOMMENDS you is what AgentReady Monitor measures. → https://ai-agent-ready.com --- A product by The Autopilot — https://the-autopilot.com