The engine behind every finding

Clir combines automated accessibility checks with browser-based analysis of rendered websites. Every confirmed issue is connected to the affected element, supporting evidence, a relevant WCAG criterion and practical guidance. Best practices and items for review are organised separately, keeping the report focused and useful.

Our engine is evaluated against established public accessibility test collections, comparable scanning tools and controlled fix-verification tests. Results are presented with their precise test scope so you can understand what the numbers demonstrate.

Finding the issue is only the beginning. Clir provides the details needed to investigate it: the affected element, the relevant requirement, why the issue matters and recommended corrections.

• 98.0%Equivalent-rule precisionIBM shared-rule evaluation · 2 Nov 2026
• 91.5%Shared-rule detectionIBM test targets · 2 Nov 2026
• 1/250ACT false positivesPassing or inapplicable development examples

Equivalent-rule precision is measured on IBM-labelled targets for which an equivalent check is available.

High-confidence results, with clear next steps

Confirmed WCAG findings identify issues supported by automated technical evidence. Best-practice recommendations help teams improve beyond individual confirmed failures. Needs Review highlights questions worth checking, with context that makes follow-up easier.

False-positive testing

Clir is tested against positive and negative examples, including accessible reference components. We track incorrect findings and refine checks while preserving genuine detections. Results shown here are from the labelled benchmark cases described below.

The comparisons below show confirmed detections on defined public test sets under comparable browser and scoring conditions. The shared-rule comparison highlights detection on the same accessibility checks across engines.

Each chart names its dataset and evaluation scope. Our latest same-page IBM common-rule comparison measures CLIR's detection against axe-core and IBM Equal Access on the same targets.

Rendered pages, not just source code

Clir evaluates pages as they appear in a browser, including content created as a page loads. Checks cover accessible names, forms, images, ARIA relationships, contrast, headings, page structure and supported keyboard and focus behaviour.

Language-aware accessibility review

Clir checks for language mismatches between declared page languages and the content of pages and marked multilingual passages. Language review findings include supporting details and practical next steps.

One page or a defined multi-page assessment

Start with a free scan of a specific page, or select a set of pages for a larger assessment. For multi-page work, Clir helps discover candidate pages, records the confirmed assessment scope and organises results across the pages tested. Only the pages confirmed as part of the selected scope are included in the assessment.

From issue to correction

Findings include affected elements, impact, relevant standards and practical remediation guidance. The Accessibility Report and Fix Pack organise this evidence into work developers and agencies can take forward.

After changes are made, Clir rechecks the relevant pages and tests the original findings again. The result tracks what now passes, what remains detected and what has changed, providing a clear record of progress.

• 100%Genuine fixes credited61 controlled fixes
• 0False resolutions473 controlled non-fix tests

Actionable reports and a lasting record

The Fix Pack turns findings into corrections and verifiable outcomes. The Closeout brings the initial assessment and subsequent rechecks together into a dated record that teams can share and revisit.

Precise verification outcomes

A finding is recorded as resolved when the original check passes again on the matched element. Changes to or removal of an element are tracked separately, making the history of the work clear.

Public, traceable test material

CLIR uses W3C ACT examples, IBM Equal Access test pages, GOV.UK accessibility test material and controlled before-and-after cases. Every displayed score is tied to a named dataset, a defined denominator and a particular engine version.

Comparable results

Engines are evaluated on the same relevant pages with documented browser settings and matching rules. A confirmed detection is credited using the defined element and WCAG criterion. Shared-rule comparisons are distinguished from IBM-specific policy checks.

Finding quality

Equivalent-rule precision measures agreement with labelled expected outcomes for corresponding checks. False-positive tests use examples expected to pass or be inapplicable. Fix verification tests check both genuine corrections and changes that should not count as fixes.

Evidence and evaluation scope

Each result is linked to its evaluated test material, scoring method and engine version, making the comparisons traceable and easy to interpret.

Detailed results

The complete scorecard preserves the counts behind the headline percentages and identifies each evaluation set.

CLIR connects confirmed technical findings with applicable WCAG 2.2 Level A and AA success criteria. Optional AAA checks are labelled separately. Findings may also include context for ADA, Section 508 and the European Accessibility Act / EN 301 549 where relevant mappings are supported.

This gives teams clear references for planning accessibility work and discussing requirements with developers, clients and stakeholders.

Each report records the pages evaluated and the findings observed in that assessment. Rechecks give teams a way to compare the same issues after changes are made.

For checks that benefit from human judgement, CLIR provides a dedicated review queue with the relevant evidence and next steps.

Privacy and sharing

Results are available through private report links, with assessment and retention details described in the Privacy Policy. Payments are handled through Stripe.