Changelog

Reports and MCP output now describe public snippet and cache directives on the analysed page: Google nosnippet, max-snippet (value preserved) and data-nosnippet (content-level, never page-wide), and Bing NOCACHE and NOARCHIVE, with their source (robots meta tag or X-Robots-Tag) and audience. They are read separately from indexability and are posture, not defects: no Recommendation or Fix Pack item, and no scoring, coverage or Methodology change.

/methodology, /limitations and llms.txt now state that inclusion in Google's generative search features (AI Overviews, AI Mode, generative features in Discover) may also depend on an owner-side Search Console setting that Existence Compiler cannot observe externally; EC reports offering to the web search channel, not inclusion in those features. Reports and MCP output now describe robots.txt exclusions of documented crawler tokens grouped by purpose (search / answer retrieval, model training, user-triggered fetch), with current tokens such as Claude-SearchBot, OAI-SearchBot, Google-Agent and Google-GeminiNotebook and the vendor caveat for user-triggered fetchers. Descriptive only: no scoring, coverage or Methodology change.

The public pages now describe the product as it measures today. The score is the EC Score (formerly the Digital Existence Score): offering to public discovery × qualitative fitness, always shown with its coverage, and not published when coverage is insufficient. /methodology is the single public source for these definitions (EC Score, coverage, CPP, lenses, Recommendations, Fix Pack, Methodology versioning, and what is not claimed). The former /examples pages, which showed a retired report model, now redirect to /example. Pricing now states that Compile Once is available. No scoring logic changed.

Added /es/puntaje-existencia-digital and /es/checklist-busqueda-ia, Spanish fixed-language server-rendered versions of the Digital Existence Score explainer and the AI Search Readiness Checklist, with Spanish metadata, hreflang pairs to their English routes, and sitemap alternates. Together with /es/capa-de-existencia-digital they form the first Spanish indexable cluster. English remains the global authority language; no scoring or report logic changed.

Published the initial skeleton of the DEL living paper at /digital-existence-layer/paper — abstract, problem statement, definition, the six components, measurement model, use cases, limitations, and update policy, with a public revision history, a how-to-cite block, and TechArticle structured data. English-only for now. Existence Compiler is described as one implementation of the DEL audit/fix model, not the definition itself.

Added /digital-existence-score, a public page explaining what the score measures, how its five dimensions plus agentic readiness form the Digital Existence Layer, why the other report surfaces are independent of the score, and how the Fix Pack acts as the implementation bridge. The report UI now shows a surface-independence legend and a connective line under the score breakdown. No scoring rules, weights, or report logic changed.

Added /digital-existence-layer and /es/capa-de-existencia-digital, the canonical definition pages for the Digital Existence Layer — a formal name for the architecture Existence Compiler has always measured. Both pages are server-rendered in a fixed language with hreflang alternates, and ship with WebPage, BreadcrumbList, FAQPage, and (on the English page) DefinedTermSet structured data covering the six components. No scoring weights or report logic changed.

Added /ai-search-readiness-checklist, a public, indexable checklist covering eight areas — AI Overview readiness, AI Mode conversational depth, information-agent monitorability, multimodal asset clarity, generative UI transformability, action/contact clarity, classic search basics, and passive trust signals. Each item explains what it is, why it matters, and what Existence Compiler checks, with BreadcrumbList, WebPage, and ItemList structured data.

Added /agents (how AI agents use Existence Compiler to turn a URL into a structured diagnostic object) and /mcp (public docs for the /api/mcp endpoint, bearer-token auth, the analyze_url, get_fix_pack, and compare_urls tools, and the sanitized output shape). Backend findings and paid Fix Pack snippets are not exposed; both pages ship with BreadcrumbList and WebPage structured data.

Added /methodology (how each surface is evaluated), /baseline (the current self-scan of existencecompiler.com as a point-in-time snapshot), and /limitations (what the tool cannot guarantee or fully observe). Each measures structural conditions, not outcomes, with BreadcrumbList and WebPage structured data.

Added /example-report, a static page explaining what a report contains — Digital Existence Score, AI & Search Visibility, Google AI & Agentic Search Readiness, and the Fix Pack — plus deeper question-style sections on the landing page.

The Fix Pack now emits targeted, copy-paste recommendations when agentic checks are weak: image alt text, FAQPage/QAPage schema, RSS/update signals, structured/generative content, conversational depth, and ContactPoint schema.

Reports now include a standalone Google AI & Agentic Search Readiness panel and a "four ways your project gets discovered" explainer covering classic search, AI Overviews, AI Mode, and agentic discovery. It measures structural readiness only.

Added structural signal extraction (image alt coverage, JSON-LD @types, RSS/Atom feeds, update dates, tables/lists, pricing/comparison/step signals) and a standalone readiness score across six checks. The headline Digital Existence Score is unchanged.

Improved text contrast across the report and marketing pages, and refined the mobile navigation strip for Pricing, FAQ, and Example.

Hardened the analysis and admin surfaces, tightened the checkout guard and refund handling paths, and added regression tests covering those flows.

Added a weighted set of structural checks — title, meta description, canonical, robots.txt, sitemap, llms.txt, JSON-LD, Open Graph, Twitter Cards, hreflang, and noindex detection — reported as a separate visibility score.

Exposed analysis capabilities over the Model Context Protocol so compatible AI agents and tools can request a structured report programmatically.