Ruleset v1.2.0 · Engine 1.2.0
AI readiness
A deterministic measurement of a public website. It is not an SEO score, not a certificate, and not a prediction about any assistant.
Website AI Readiness is not Publication Readiness. The crawl is bounded. A fact is kept only with a source URL.
97.8/100
Advanced
- Below 40Foundational
- Below 60Developing
- Below 75Moderate
- Below 90Strong
- 90 and aboveAdvanced
Diagnostic matrix
| Dimension | Weight | What is read | Evidence when present | Not applicable |
|---|---|---|---|---|
| Technical Foundation | 14 | HTTPS, response, content type, and the other technical signals. | The home page answered over HTTPS. | The site could not be fetched. |
| Business Identity | 12 | Name, contact, and organization identity the pages state. | The pages name the business. | No identity block was present to score. |
| Products & Services | 8 | Offers the pages describe. | A product or service is stated with a source URL. | The site does not offer products or services. |
| Policies & FAQs | 6 | Policy and question pages that were fetched. | A policy or FAQ page was found. | The profile has no policy surface. |
| Structured Data | 8 | JSON-LD and other structured blocks on the fetched pages. | A block parses and matches the page. | No structured block was present. |
| Machine Readability | 10 | Headings, titles, and text a parser can follow. | Heading levels and titles are usable. | The page was not HTML. |
| Crawl / Discoverability | 8 | Links, sitemap, and robots that the crawl was allowed to use. | Discovery stayed inside the budget. | Robots left nothing to fetch. |
| Canonical Consistency | 8 | Whether the same fact is stated the same way. | Titles and names agree across the sample. | A single-page sample has no second page to compare. |
| Search Readiness | 8 | Page facts plus whether this business is in the public catalog. | The catalog index is current and this host is present. | The business is absent, or the index was not tested. |
| Semantic / RAG Readiness | 8 | Whether stored public text could be cited later. RAG is not executed. | Facts have source URLs. | No citable sentence was extracted. |
| API Readiness | 3 | A public API the site itself documents. | The site links a machine-readable API. | The site does not offer an API. |
| MCP Readiness | 2 | A public MCP link the site itself documents. | The site documents an MCP link. Tools are not retested here. | The site does not offer MCP. |
| Agent Discovery Readiness | 6 | Public identity and offer signals. The external network is not started. | A public name and offer were observed. | No discovery signal was present. |
| Commerce Readiness | 6 | Offer facts the pages state. Checkout is not enabled. | An offer is described. | The site has no commerce. |
| Freshness | 4 | Dates and change signals on the fetched pages. | A date is stated. | No date was on the page. |
| Conflict Quality | 3 | Facts that disagree across the sample. | No conflict was found. | A name or price disagrees. |
| Missing Information | 2 | Core public facts the pages never state. | A name and a contact or policy path were found. | A required core fact was absent. |
| Visibility / Privacy Readiness | 10 | Whether private data was exposed, and whether a source URL was kept. | The sample stayed on public pages. | The fetch was refused. |
What the score does not do
- Not applicable means the rule had no input. It is left out. It is not a failure.
- A blocker stops publication readiness. It does not rewrite the readiness score into a rank.
- Reaching the page budget is a completed sample. Evidence stays Partial when some pages may not have been included.
- Impact stars and priority are separate from the 97.8 score.
Methodology
The assessment is deterministic. It does not call a language model and it does not invent a score. The free scan reads up to 25 public HTML pages on the same site, up to 200000 bytes on a page, and stops within 40 seconds. Ruleset v1.2.0. Engine 1.2.0. Applicable rules form the weighted score. Not applicable, not yet tested, and unknown results are left out of that score.
Dimensions
- Technical Foundation
- Business Identity
- Products & Services
- Policies & FAQs
- Structured Data
- Machine Readability
- Crawl / Discoverability
- Canonical Consistency
- Search Readiness
- Semantic / RAG Readiness
- API Readiness
- MCP Readiness
- Agent Discovery Readiness
- Commerce Readiness
- Freshness
- Conflict Quality
- Missing Information
- Visibility / Privacy Readiness
A business profile changes which rules apply. A site with no commerce signals does not lose points for a missing checkout. A site that never links an API does not lose points for API readiness.
Scores and evidence
Ruleset v1.2.0 labels are Foundational work required below 25, Low readiness below 50, Partial readiness below 75, Good readiness with improvements below 90, and Strong technical readiness at 90 or above. Stored v1.1.0 reports keep the labels they were given. Evidence quality is separate from the score. A high score does not clear a blocker.
Impact stars describe how much a finding can matter. Priority describes what to do first. They are not the score.
Publication readiness
Publication readiness is a separate label. Ready requires a strong score, complete evidence, and no blocker. The free scan never publishes the website.
Older reports
Stored jobs that have no ruleset version stay labeled Basic Technical Readiness. Those older checks are not the current scan, and their scores are not compared with v1.2.0.