Public evidence
Edikka accessibility audit and evidence
Method, current results, historical status and limitations remain separate.
On August 26, 2026, Edikka documented all 36 controls in its Accessibility Foundation across eight representative journeys. Each result keeps its status and evidence. This targeted self-assessment does not replace an exhaustive RGAA audit or third-party certification.
- 36 documented controls
- 8 representative journeys
- 4 statuses kept distinct
- JSON, XLSX and Markdown evidence
Results without an invented conformance score
“Further testing required” means an attempt exists but the evidence is insufficient. “Out of scope” requires a written reason. Neither status is converted into a pass.
- Pass
- 24
- Fail
- 0
- Further testing
- 7
- Out of scope
- 5
Claim boundary: this targeted self-assessment is neither an exhaustive RGAA audit nor a certification. It documents the available evidence and keeps insufficient evidence visible.
Eight representative journeys
The 36 controls and their evidence
Control definitions below are translated in the bilingual source method. Recorded observations remain available in the canonical assessment JSON; evidence files are language-independent technical artefacts.
A11Y01Identifiable page titlePass
User barrier: The user cannot tell which page is open.
Acceptance rule: The browser title describes the page and distinguishes it from other pages.
Test method: Compare the title element, H1 and visible topic.
Reference: WCAG 2.4.2 (A) · 8 · Éléments obligatoires
Evidence:
A11Y02Declared primary languagePass
User barrier: Speech output pronounces content incorrectly.
Acceptance rule: The html lang attribute matches the actual primary language.
Test method: Inspect html[lang] and read a passage with assistive technology.
Reference: WCAG 3.1.1 (A) · 8 · Éléments obligatoires
Evidence:
A11Y03Coherent heading hierarchyPass
User barrier: Sections become difficult to scan or understand.
Acceptance rule: One H1 describes the topic; H2/H3 reflect content nesting and are not decorative.
Test method: Read the heading outline without visual styling.
Reference: WCAG 1.3.1 (A); 2.4.6 (AA) · 9 · Structuration
Evidence:
A11Y04Landmarks and direct access to contentPass
User barrier: Repeated navigation must be traversed on every page.
Acceptance rule: Main content is inside main; a working bypass link reaches a visible or focusable target.
Test method: Tab from the start, then inspect landmarks.
Reference: WCAG 1.3.1 (A); 2.4.1 (A) · 9 · Structuration; 12 · Navigation
Evidence:
A11Y05Alternatives for informative imagesFurther testing required
User barrier: Information conveyed by an image disappears.
Acceptance rule: Each informative image has an equivalent alternative; decorative images can be ignored.
Test method: Hide images and compare the remaining information; inspect alt and role.
Reference: WCAG 1.1.1 (A) · 1 · Images
Evidence:
A11Y06Names for functional images and iconsPass
User barrier: A graphical action is silent or ambiguous.
Acceptance rule: The accessible name describes the action or destination, not the icon appearance.
Test method: Inspect the computed name of graphical controls.
Reference: WCAG 1.1.1 (A); 4.1.2 (A) · 1 · Images; 7 · Scripts
Evidence:
A11Y07Links understandable in contextFurther testing required
User barrier: The destination of a link must be guessed.
Acceptance rule: The label and programmatic context identify the destination or function.
Test method: List links outside the layout and check vague repetitions.
Reference: WCAG 2.4.4 (A) · 6 · Liens
Evidence:
A11Y08Instructions independent of shape and colourFurther testing required
User barrier: Instructions become unusable without full visual perception.
Acceptance rule: No essential instruction relies only on colour, position, shape or sound.
Test method: Review instructions after removing sensory cues.
Reference: WCAG 1.3.3 (A); 1.4.1 (A) · 3 · Couleurs; 13 · Consultation
Evidence:
A11Y09Text contrastPass
User barrier: Text becomes unreadable for many users.
Acceptance rule: Contrast reaches 4.5:1 for normal text and 3:1 for large text, subject to WCAG exceptions.
Test method: Measure colour pairs in their real states.
Reference: WCAG 1.4.3 (AA) · 3 · Couleurs
Evidence:
A11Y10Contrast of components and meaningful statesPass
User barrier: A field, focus indicator or state cannot be perceived.
Acceptance rule: Necessary visual boundaries and states reach 3:1 against adjacent colours, subject to WCAG exceptions.
Test method: Measure fields, buttons, focus, graphics and active states.
Reference: WCAG 1.4.11 (AA) · 3 · Couleurs; 10 · Présentation
Evidence:
A11Y11Text zoom and custom spacingFurther testing required
User barrier: Content overlaps or disappears when reading settings are adapted.
Acceptance rule: At 200% zoom and WCAG text spacing, text and functions remain available without loss.
Test method: Apply zoom and a text-spacing stylesheet; replay the journey.
Reference: WCAG 1.4.4 (AA); 1.4.12 (AA) · 10 · Présentation
Evidence:
A11Y12Reflow at 320 CSS pixelsPass
User barrier: Two-dimensional reading or horizontal scrolling is imposed.
Acceptance rule: Content and actions remain usable at 320 CSS px without loss or global horizontal scrolling, subject to exceptions.
Test method: Test at 320 px and 400% on a suitable viewport.
Reference: WCAG 1.4.10 (AA) · 10 · Présentation
Evidence:
A11Y13All functions available by keyboardPass
User barrier: A person without a mouse cannot complete the journey.
Acceptance rule: Every essential action can be completed by keyboard without pointer-dependent gestures.
Test method: Replay the journey with Tab, Shift+Tab, Enter, Space and arrows as appropriate.
Reference: WCAG 2.1.1 (A) · 7 · Scripts; 12 · Navigation
Evidence:
A11Y14No keyboard trapPass
User barrier: Focus enters a component and cannot leave it.
Acceptance rule: Focus can leave every component through a standard or documented method.
Test method: Enter and leave menus, dialogs, players and widgets by keyboard.
Reference: WCAG 2.1.2 (A) · 7 · Scripts; 12 · Navigation
Evidence:
A11Y15Visible and unobscured focusPass
User barrier: The user loses their position or focus is covered.
Acceptance rule: Every focused control is perceptible and not entirely hidden by fixed content.
Test method: Tab through the journey, including mobile and sticky headers.
Reference: WCAG 2.4.7 (AA); 2.4.11 (AA) · 10 · Présentation; 12 · Navigation
Evidence:
A11Y16Logical focus orderPass
User barrier: Reading and actions follow an incoherent order.
Acceptance rule: Sequential focus preserves meaning and operation; no positive tabindex forces it artificially.
Test method: Compare DOM order, visual order and tab sequence.
Reference: WCAG 2.4.3 (A) · 12 · Navigation
Evidence:
A11Y17Accessible name aligned with visible labelPass
User barrier: Voice control cannot find the displayed control.
Acceptance rule: The accessible name contains the visible action text in a useful matching order.
Test method: Compare the visible label and computed name.
Reference: WCAG 2.5.3 (A) · 7 · Scripts
Evidence:
A11Y18Programmatic name, role, value and statePass
User barrier: The component is announced without a function or state.
Acceptance rule: Each component exposes a name, role and, where needed, an updated value or state.
Test method: Inspect the accessibility tree before and after interaction.
Reference: WCAG 4.1.2 (A) · 7 · Scripts
Evidence:
A11Y19Focus management for dialogs and panelsPass
User barrier: Context changes without orientation or background content remains active.
Acceptance rule: On open, focus enters the component; it stays when required; on close, it returns to a logical point.
Test method: Open, traverse and close by button and Escape, then check focus return.
Reference: WCAG 2.4.3 (A); 2.1.2 (A); 4.1.2 (A) · 7 · Scripts; 12 · Navigation
Evidence:
A11Y20Announced status messagesFurther testing required
User barrier: Success, error or update feedback remains invisible to speech output.
Acceptance rule: Important messages are exposed without unnecessarily moving focus.
Test method: Trigger search, add, submit and error states, then listen to announcements.
Reference: WCAG 4.1.3 (AA) · 7 · Scripts; 11 · Formulaires
Evidence:
A11Y21Explicit label for every fieldPass
User barrier: The expected data must be guessed.
Acceptance rule: Each field has a persistent, correctly associated and sufficiently precise label.
Test method: Hide placeholders, then inspect label/control associations.
Reference: WCAG 1.3.1 (A); 3.3.2 (A) · 11 · Formulaires
Evidence:
A11Y22Identifiable purpose for personal-data fieldsPass
User barrier: Input assistance and autocomplete cannot work correctly.
Acceptance rule: Common personal-data fields use a valid autocomplete token when their purpose is known.
Test method: Inspect name, type and autocomplete for relevant fields.
Reference: WCAG 1.3.5 (AA) · 11 · Formulaires
Evidence:
A11Y23Errors identified and linked to fieldsFurther testing required
User barrier: The user knows submission failed but not where or why.
Acceptance rule: The error is named, located, linked to the field and announced; correct data is retained.
Test method: Submit empty, then with several controlled errors.
Reference: WCAG 3.3.1 (A); 3.3.3 (AA) · 11 · Formulaires
Evidence:
A11Y24Prevention and confirmation for sensitive actionsOut of scope
User barrier: A legal, financial or irreversible action is confirmed by mistake.
Acceptance rule: The user can review, correct or confirm before finalisation and receives usable confirmation.
Test method: Replay the critical journey with a deliberate error and correction.
Reference: WCAG 3.3.4 (AA) · 11 · Formulaires
Evidence:
A11Y25Captions for prerecorded videoOut of scope
User barrier: Audio content is inaccessible to deaf or hard-of-hearing users.
Acceptance rule: Speech and meaningful sounds are available in synchronised, reviewed captions.
Test method: Play the video without sound and compare it with the captions.
Reference: WCAG 1.2.2 (A) · 4 · Multimédia
Evidence:
A11Y26Alternative for essential visual or audio contentPass
User barrier: A demonstration, chart or meaningful sound has no equivalent.
Acceptance rule: A transcript, audio description or text alternative conveys the necessary information for the medium.
Test method: Identify what is lost when image or sound is removed.
Reference: WCAG 1.2.3 (A); 1.2.5 (AA) · 4 · Multimédia
Evidence:
A11Y27Control of autoplay, motion and timingPass
User barrier: Content disrupts reading or expires before the action is complete.
Acceptance rule: Automatic audio can be stopped; long animations can be paused; time limits can be controlled, subject to exceptions.
Test method: Load the page, wait five seconds, and test pause, stop and extension.
Reference: WCAG 1.4.2 (A); 2.2.1 (A); 2.2.2 (A) · 4 · Multimédia; 13 · Consultation
Evidence:
A11Y28Safe flashing and controllable motionPass
User barrier: Flashing or automatically moving content causes discomfort, loss of focus or neurological risk.
Acceptance rule: No content flashes more than three times per second and any automatic animation lasting more than five seconds can be paused, stopped or hidden.
Test method: Measure flashing, then verify pause, stop and hide controls for each relevant automatic animation.
Reference: WCAG 2.2.2 (A); 2.3.1 (A) · 13 · Consultation
Evidence:
A11Y29Orientation not restrictedPass
User barrier: The service becomes unusable on a device fixed in one orientation.
Acceptance rule: Content works in portrait and landscape unless orientation is essential.
Test method: Replay the journey in both orientations.
Reference: WCAG 1.3.4 (AA) · 13 · Consultation
Evidence:
A11Y30Sufficiently large touch targetsPass
User barrier: An action is difficult to trigger accurately on mobile.
Acceptance rule: Targets reach 24 × 24 CSS px or meet an exception in WCAG 2.5.8.
Test method: Measure the box and spacing of small targets at mobile viewport.
Reference: WCAG 2.5.8 (AA) · 10 · Présentation; 13 · Consultation
Evidence:
A11Y31Consistent navigation and componentsPass
User barrier: The same element changes name, position or behaviour without reason.
Acceptance rule: Repeated navigation keeps a consistent order and identical components are identified consistently.
Test method: Compare at least three pages and two instances of each critical component.
Reference: WCAG 3.2.3 (AA); 3.2.4 (AA) · 12 · Navigation
Evidence:
A11Y32Coherent programmatic reading orderFurther testing required
User barrier: Visual reordering disconnects reading, meaning and action.
Acceptance rule: At each viewport, DOM order and assistive-technology output preserve the meaning of content and instructions.
Test method: Compare visual order, DOM order, keyboard navigation and screen-reader output on the same state.
Reference: WCAG 1.3.2 (A) · 9 · Structuration; 10 · Présentation
Evidence:
A11Y33Alternative to dragging movementsOut of scope
User barrier: An action requires precise drag-and-drop that a person cannot perform.
Acceptance rule: Any function based on dragging can also be completed with a single-pointer action, unless dragging is essential.
Test method: Identify drag-and-drop, carousel and reordering interactions, then complete each action without holding and moving the pointer.
Reference: WCAG 2.5.7 (AA) · 7 · Scripts — critère WCAG 2.2 sans équivalent direct dans le RGAA 4.1.2
Evidence:
A11Y34Consistent help across pagesPass
User barrier: The way to obtain help changes position or order and becomes difficult to find.
Acceptance rule: When a help mechanism repeats across pages, it keeps the same relative order unless the user initiated the change.
Test method: Compare at least three pages offering contact, assistant, FAQ or contextual help at the same viewports.
Reference: WCAG 3.2.6 (A) · 12 · Navigation — critère WCAG 2.2 sans équivalent direct dans le RGAA 4.1.2
Evidence:
A11Y35Redundant entry avoidedOut of scope
User barrier: Information already supplied must be entered again in the same process.
Acceptance rule: Previously supplied information is auto-populated or available for selection, except for security, necessity or invalid-data exceptions.
Test method: Replay a multi-step journey using the same contact details, addresses or business information.
Reference: WCAG 3.3.7 (A) · 11 · Formulaires — critère WCAG 2.2 sans équivalent direct dans le RGAA 4.1.2
Evidence:
A11Y36Accessible authenticationOut of scope
User barrier: Signing in requires a cognitive test, memorisation or transcription without an alternative.
Acceptance rule: Authentication does not require a cognitive-function test, or provides an alternative, assistance mechanism, or object/personal-content recognition covered by WCAG exceptions.
Test method: Test sign-in, account recovery and multi-factor authentication with a password manager, copy and paste, and available assistance.
Reference: WCAG 3.3.8 (AA) · 11 · Formulaires — critère WCAG 2.2 sans équivalent direct dans le RGAA 4.1.2
Evidence:
Internal RGAA audit — June 25, 2026
The historical internal audit reported 76 conformant criteria, three non-conformities and 27 non-applicable criteria: 96% of applicable criteria met. This dated finding remains separate from the current 36-control self-assessment and is not presented as third-party certification.
Criterion 4.1 remains a historical non-conformity. An August 26 correction addendum documents the audio-template change and the production counter-tests still required.