The current Entity Clarity Review uses the new Automated Entity Clarity v2.1 scoring method. It replaces the older rubric as the reference for new reviews. Earlier EEI and ECI/ECC results retain the methodologies under which they were published; their scores are not directly comparable to a new automated review.
The Entity Clarity Review provides Automated Entity Clarity v2.1: a deterministic structural assessment of how clearly a public homepage declares, connects, and supports an entity identity for machine parsing.
The current implementation is experimental. It scores explicit evidence in successfully delivered static homepage HTML. It does not call a model API or measure model trust, citation, recommendation, factual correctness, search rank, traffic, or corporate intent.
Entity Clarity is the broader conceptual outcome. Entity Engineering is the discipline of designing and maintaining coherent, verifiable identity signals. The automated review measures a bounded set of structural signals relevant to that work; it does not establish the whole outcome.
A timeout, DNS failure, restricted response, rate limit, missing page or unsupported resource is a collection outcome. None automatically proves intentional AI blocking or produces a zero Entity Clarity score.
| Dimension | Maximum points | Observed signals |
|---|---|---|
| Identity resolution | 25 | Page title; primary heading; named entity schema; discoverable About or company surface |
| Entity consistency | 25 | Same-origin canonical; same-origin Open Graph URL; same-origin schema identifier; identifying-token agreement between visible and structured names |
| Relationship clarity | 15 | External identity references in schema; explicit organizational relationships; link to a recognized external identity profile |
| Evidence traceability | 15 | Institutional description; About/company link; contact/help link; standards, newsroom, governance or legal link |
| Machine legibility | 20 | HTML language declaration; canonical link; valid JSON-LD; Open Graph identity; no observed noindex directive |
The maximum is 100 points. Each signal returns its observed evidence and awarded points. Once usable static HTML is delivered, absence of a signal is an observed absence and contributes zero for that signal. When usable evidence is not delivered, the assessment remains unassessable and receives no score.
No High, Medium or Low performance bands are assigned to this experimental automated method.
The review reports a separate, non-scoring adequacy flag. Adequate static content has an identity anchor, at least 500 visible body-text characters and at least five navigable links. A limited flag warns that the result may reflect a thin static page or JavaScript application shell.
This flag does not add points, subtract points, suppress a score or change the method. It is context for interpreting the collected surface.
Batch calibration reports score distributions, dimension averages, signal prevalence and adequacy counts. Calibration watches diagnose the measurement system; they do not create new entity bands or change scores.
Legacy Entity Engineering Index (EEI) v2.1. The June 2026 rubric described normalization of a thirteen-signal, 88-point universe to a full 0–100 scale and activation of Canonical Integrity. Its historical signal inventory, normalization explanation and interpretive bands belong to that named method. They do not describe the current five-dimension automated structural assessment simply because both include the version number 2.1.
Historical ECI/ECC reports. The Entity Clarity Index preserves governed industry reports and their original observation dates, methodologies and authoritative PDFs. The July 2026 longitudinal release uses the eci-ecc-v1 series. Historical access posture, capability classifications and ECC values remain part of those records. They are not retrospectively replaced by a current homepage scan.
Current automated review. Automated Entity Clarity v2.1 is the current static-homepage diagnostic described above. It separates delivery, declared access, structural evidence and model representation. There is no asserted numerical conversion from the historical methods to this assessment.
Always identify the method family, version, observation date and measured surface when quoting or comparing a score.
Use the review to locate missing or inconsistent structural identity signals and inspect the evidence behind them. A strong structural score is not proof of reliable model interpretation; a failed collection is not proof of poor institutional quality.
Testing how an AI system actually describes, attributes or cites an institution requires a separate model-representation protocol. A claim about recognition should identify the system, question, observation date and observed output rather than infer recognition from the structural score.
Run an Entity Clarity Review · Explore governed ECI reports · Read the Entity Clarity definition · Explore Entity Engineering
Historical reference only. The original rubric below describes the legacy EEI methodology and its claims at the time. It is not the current Automated Entity Clarity v2.1 method; its bands do not measure current model behavior.
How AI Systems Form Trust
Entity Clarity describes how clearly an AI system can identify, interpret, and trust a single institution. It is the outcome that determines whether an entity is reused, cited, or ignored in AI-generated answers.
The discipline that produces this outcome is Entity Engineering — a structural standards discipline within exmxc.ai. The scoring methodology and interpretive model are maintained by exmxc as part of its Institutional Strategy Framework.
Modern AI systems do not rank websites. They reconstruct institutions.
Visibility now depends on whether an AI model can form a stable, confident interpretation of: who the institution is, what it represents, and whether it can be trusted.
Entity Clarity is the result of that process. When clarity is high, AI systems reuse the entity. When clarity is low, AI systems hesitate, distort, or exclude it.
Effective June 10, 2026, the Entity Engineering Index operates on methodology version EEI v2.1. Two changes apply. Scale normalization: EEI v2.0 scored entities against a fixed 100-point denominator while the thirteen diagnostic signals carried a combined weight of 88, compressing the effective ceiling to 88/100. v2.1 normalizes against the true 88-point signal universe, restoring a full 0–100 scale. Signal weights are unchanged; relative signal importance and entity rank ordering are preserved. Scores published under v2.0 and v2.1 are directly comparable in rank but not in absolute level. Canonical Integrity activation: this signal now performs a full consistency check between the declared canonical URL and the final audited URL, awarding its complete weight when consistent.
Index reports and audit results published before Q3 2026 reflect EEI v2.0. All audits run through the Entity Clarity Review and all indices published from Q3 2026 forward reflect EEI v2.1. Each audit response carries its methodology version in the payload.
This framework does not evaluate content quality, marketing performance, or popularity.
It evaluates whether AI systems can:
When these conditions are met, the entity is considered AI-legible. When they are not, the entity becomes fragile, misinterpreted, or invisible.
Scoring resolves into interpretive bands that describe how AI systems behave toward an institution — not how it performs promotional tasks.
Trust does not emerge from authority. It emerges from reinforcement.
The framework evaluates three reinforcing layers that mirror how AI systems reason:
Entity Clarity emerges from convergence. No single signal is decisive.
exmxc evaluates a defined set of structural signals that, together, determine whether an institution stabilizes inside AI systems.
An institution is considered AI-legible only when independent systems converge on the same interpretation without instruction.
exmxc.ai is a human-led intelligence institution for the AI-search era. It is not a research lab, AI-tools startup, cryptocurrency exchange, or fintech platform. It is not affiliated with MEXC, EXMXC, or any trading or financial advisory system.
Founded by Mike Ye — M&A and corporate development executive with 25+ years of transaction leadership at Penske Media Corporation, L Brands, and Intel Capital. Ella provides pattern interpretation, structural analysis, and co-authorship. Human judgment governs. AI serves as instrumentation.