`, no skipped levels), live-region presence on every async surface, and an **exhaustive WCAG contrast measurement of every design-token colour pairing** (41 pairings × 2 themes) resolved against the rendered pixels. | | **What "automated AT" means here (and its limit)** | The deciding evidence for the *programmatic* success criteria (Name/Role/Value, Status Messages, Info & Relationships) is **what the product EXPOSES to assistive technology through the platform accessibility API** — and that is exactly what the harness verifies directly, in a real browser, across the whole surface. **A human screen-reader walkthrough — an operator listening to NVDA, JAWS or VoiceOver complete tasks — has NOT been performed** (and cannot be in the project's headless Linux CI). For the SC-level rows this API-level verification is sufficient and they are reported **Supports**. For the holistic **Functional-Performance** statements about *non-visual use as a whole* (508 **302.1** / EN **4.2.1**), the building blocks are all verified, but a **human screen-reader spot-test is still recommended** before relying on this ACR for a specific high-stakes regulated/gov procurement; those rows say so explicitly. | | **Notes** | The EAA's harmonised standard, **EN 301 549 V3.2.1**, normatively references **WCAG 2.1 AA** — so the Chapter 9 table below is the 2.1 baseline that carries the presumption of conformance. The six new-in-2.2 A/AA success criteria (2.4.11, 2.5.7, 2.5.8, 3.2.6, 3.3.7, 3.3.8) are reported in a separate section ahead of EN 301 549 V4.1.1, based on actual testing. SC 4.1.1 Parsing was removed in WCAG 2.2 and is not reported. | > **Honesty statement.** This is a factual self-assessment. Every "Supports" is backed by code + axe + keyboard + **platform-accessibility-API** evidence cited in Remarks; conformance is asserted only for what was exercised. The method is explicit about its one limit (above): the programmatic exposure that AT consumes is verified directly and automatically; a human screen-reader walkthrough was not run, so the two holistic functional-performance rows carry a residual human-spot-test recommendation rather than overclaiming. Re-issue this ACR (a) if a human/third-party NVDA·JAWS·VoiceOver pass is later commissioned (fold its result in, drop the residual note), (b) when EN 301 549 V4.1.1 becomes harmonised, and (c) as later business views ship (the AT harness re-runs per view and must stay green — it is wired as `pnpm -C web at:gate`). ## Conformance terms (ITI vocabulary) | Term | Meaning | |---|---| | **Supports** | At least one method meets the criterion without known defects (or meets with equivalent facilitation). | | **Partially Supports** | Some functionality does not meet the criterion. | | **Does Not Support** | The majority of functionality does not meet the criterion. | | **Not Applicable** | The criterion is not relevant to the product. | ## Applicable standards / guidelines | Standard / Guideline | In report | |---|---| | WCAG 2.1 (A, AA) — via EN 301 549 Ch. 9 | Yes (primary, legal baseline) | | WCAG 2.2 (A, AA) — new SC only, tested | Yes (see Notes) | | Revised Section 508 (2017, corrected 2018) — Ch. 3 FPC, Ch. 5 Software, Ch. 6 Support Documentation & Services | Yes (INT edition; WCAG incorporated by reference per E205.4/E207.2) | | Revised Section 508 — Ch. 4 Hardware | No — Not Applicable (rationale below) | | EN 301 549 V3.2.1 Ch. 4 Functional Performance Statements | Yes | | EN 301 549 V3.2.1 Ch. 9 Web | Yes (primary) | | EN 301 549 V3.2.1 Ch. 12 Documentation & support | Yes | | EN 301 549 V3.2.1 Ch. 5 / 6 / 7 / 10 / 11 / 13 | No — Not Applicable (rationale below) | --- ## EN 301 549 — Chapter 9: Web (WCAG 2.1 A + AA) | Criteria | Level | Conformance | Remarks and explanations | |---|---|---|---| | 1.1.1 Non-text Content | A | Supports | Icons are decorative (`aria-hidden`) beside text labels; icon-only controls carry `aria-label`. Charts expose a screen-reader summary via the `AccessibleChart` primitive (`role="img"` + description) and a data-table alternative. | | 1.2.1–1.2.5 Time-based media | A/AA | Not Applicable | The console plays no audio or video / time-based media. | | 1.3.1 Info and Relationships | A | Supports | Semantic landmarks (`main`, `header`, `nav`, `aside`), `DataTable` as `role="grid"` with row/col indices, `TreeGrid` as `role="treegrid"` with `aria-level`/`aria-expanded`, forms via the `Field` label association. | | 1.3.2 Meaningful Sequence | A | Supports | DOM order follows visual order; flex layout, no positional reordering that changes meaning. | | 1.3.3 Sensory Characteristics | A | Supports | Instructions never rely on shape/position alone; status uses label + icon, not "the green one". | | 1.3.4 Orientation | AA | Supports | Responsive; not locked to an orientation. | | 1.3.5 Identify Input Purpose | AA | Supports | Login/setup inputs use correct `autocomplete` tokens (`username`, `current-password`, `new-password`). | | 1.4.1 Use of Color | A | Supports | Status/mode never encoded by colour alone (badges carry text; confidence uses a distinct axis + label; `AccessibleChart` adds a non-colour data-table view). | | 1.4.2 Audio Control | A | Not Applicable | No auto-playing audio. | | 1.4.3 Contrast (Minimum) | AA | Supports | **Exhaustive** token-pairing audit (Rev 1.2): every semantic text pairing measured against rendered pixels in both themes — all ≥ 4.5:1 (large/UI ≥ 3:1). The audit found and fixed real misses (dark `muted-foreground` on tinted soft fills was 4.1–4.4; `danger` on `danger-soft` 4.41; the third-party React-Flow attribution link #999 was restyled to a token). axe `color-contrast` is clean across all 33 views in both themes. | | 1.4.4 Resize Text | AA | Supports | rem/em units; zoom to 200% reflows without loss. | | 1.4.5 Images of Text | AA | Supports | Text is real text (self-hosted variable fonts); no images of text. | | 1.4.10 Reflow | AA | Partially Supports | Layout reflows to small viewports (mobile nav, responsive shell); the access **graph** is desktop-only with a documented text fallback on small screens (an intentional equivalent, not a reflow of the canvas). | | 1.4.11 Non-text Contrast | AA | Supports | The information **required to identify a control and its state** meets 3:1: the focus ring (`ring-ring`) is ≥ 3:1 against adjacent surfaces in both themes (measured), and chart/graph data strokes use the solid semantic tokens (≥ 3:1). The *resting* hairline borders (`border-strong`, the thin semantic `-line` rules) are intentionally low-contrast (1.3–1.9:1) but are **not the sole identifier** — a control is identified by its fill + label + the ≥ 3:1 focus indicator, and the `-line` rules are decorative reinforcement beside text/icons (1.4.11 Understanding). | | 1.4.12 Text Spacing | AA | Supports | No fixed line-height/letter-spacing traps; content tolerates user spacing overrides. | | 1.4.13 Content on Hover or Focus | AA | Supports | Tooltips/popovers are Radix (dismissable, hoverable, persistent); the sidebar tooltip is keyboard-reachable. | | 2.1.1 Keyboard | A | Supports | Every interactive element is keyboard-operable: grid/treegrid arrow navigation, combobox, dialogs, nav, ⌘K palette, graph zoom/pan buttons + arrow-key pan. | | 2.1.2 No Keyboard Trap | A | Supports | Radix dialog/sheet/popover trap focus only while open and release on Escape with focus return (verified `dialog.test.tsx`). | | 2.1.4 Character Key Shortcuts | A | Supports | The only single-key affordance (⌘K) is a modifier combo, not a lone character. | | 2.2.1 / 2.2.2 Timing / Pause | A | Partially Supports | No session time limits imposed by the UI. The one looping animation (live-attribution pulse) honours `prefers-reduced-motion`; a user-facing pause control for it is not provided (it is decorative and stoppable via OS reduced-motion). | | 2.3.1 Three Flashes | A | Supports | No flashing content. | | 2.4.1 Bypass Blocks | A | Supports | A "Skip to content" link targets `#main-content` as the first focusable element (`app-layout.tsx`). | | 2.4.2 Page Titled | A | Supports | `