Manifest position 699 · CrUX rank bucket 1000
https://www.tax.service.gov.uk
Coverage complete
The requested URL redirects to the public HMRC organisation page on GOV.UK; the audit followed that entry point into representative HMRC information and unauthenticated support routes.
Attempts
1 / 3
Judged checks
58 / 58
Blocked
0
Not run
0
This report has no published overall score. Blocked and not-run checks are not passes. A coverage-complete report means every check has a judged outcome; it does not mean every check passed.
All 58 atomic check outcomes
| Principle / check | Status | Confidence | Evidence or reason |
|---|---|---|---|
respect-user-preferencesrespects-color-scheme | issues | high | The prefers-color-scheme: dark screenshot remains the same white and pale-blue light palette as the default capture; computed html color-scheme is normal. F1 medium: The service does not adapt its colour scheme to the user’s dark-mode preference. |
respect-user-preferencesrespects-reduced-motion | pass | high | Reduced-motion probe reported zero active animations. |
respect-user-preferencesrespects-contrast | pass | high | Forced-colors capture retains legible text, links, button boundaries and focus affordances. |
implement-natural-interactionsview-transitions | pass | high | No active animations appeared under reduced motion; CSS exposes a root view-transition name and the sampled pages use ordinary native scrolling/navigation without custom pointer handlers. |
implement-natural-interactionsscroll-driven-animations | not-applicable | high | No scroll-linked animation is present on the representative pages. |
implement-natural-interactionsphysical-gestures | not-applicable | high | No gesture-driven control is present in the representative journeys. |
provide-guided-navigationscroll-state-aware-chrome | pass | high | Breadcrumbs, contents links, descriptive route labels, back/skip links and focused chat state guide location and attention across the tested journey. |
provide-guided-navigationanchored-positioning | not-applicable | high | No tooltip, popover, or floating menu requiring anchored positioning appears in the tested paths. |
provide-guided-navigationdirects-attention | pass | high | Breadcrumbs, contents links, descriptive route labels, back/skip links and focused chat state guide location and attention across the tested journey. |
maximize-content-reduce-noiseno-intrusive-interruptions | pass | high | Screenshots show content-first GOV.UK layouts; cookie choices are visible, equally weighted and dismissible rather than modal, and no load-time overlay obscures the document. |
maximize-content-reduce-noisesemantic-dismissible-primitives | not-applicable | high | No modal, popover, custom select, or disclosure is used in the tested core path; the cookie choice is an in-flow semantic region. |
maximize-content-reduce-noisereduced-chrome | pass | high | Screenshots show content-first GOV.UK layouts; cookie choices are visible, equally weighted and dismissible rather than modal, and no load-time overlay obscures the document. |
adapt-to-the-form-factorresponsive-no-horizontal-scroll | pass | high | All four mobile layouts reported 360px scrollWidth for a 360px viewport with zero overflow; GOV.UK focus styling and @container rules were observed. |
adapt-to-the-form-factorcomponent-level-responsiveness | pass | high | All four mobile layouts reported 360px scrollWidth for a 360px viewport with zero overflow; GOV.UK focus styling and @container rules were observed. |
adapt-to-the-form-factorinput-modality-aware | pass | high | All four mobile layouts reported 360px scrollWidth for a 360px viewport with zero overflow; GOV.UK focus styling and @container rules were observed. |
support-core-task-successclear-purpose-and-primary-action | pass | high | The landing states HMRC’s purpose, links directly to sign-in and tax tasks, and the tested Ask HMRC > Self Assessment journey reached a labelled chat input, status message, Send and End chat controls. |
support-core-task-successprimary-flow-completion | pass | high | The landing states HMRC’s purpose, links directly to sign-in and tax tasks, and the tested Ask HMRC > Self Assessment journey reached a labelled chat input, status message, Send and End chat controls. |
support-core-task-successclear-system-state-and-recovery | pass | high | The landing states HMRC’s purpose, links directly to sign-in and tax tasks, and the tested Ask HMRC > Self Assessment journey reached a labelled chat input, status message, Send and End chat controls. |
be-fast-and-stablegood-core-web-vitals | issues | high | Lighthouse measured landing LCP 4.22s and CLS 0.556; the assistant measured LCP 5.07s. The trace corroborated a responsive main thread (0ms TBT), so loading and layout stability are the main concerns. F2 high: Key pages miss the good LCP bar, and the organisation landing page has severe lab layout shift. |
be-fast-and-stablevisual-stability | pass | high | Trace, HAR, layout and Lighthouse evidence directly measured loading, stability, main-thread and delivery behaviour across the landing and assistant routes. |
be-fast-and-stableefficient-main-thread | pass | high | Trace, HAR, layout and Lighthouse evidence directly measured loading, stability, main-thread and delivery behaviour across the landing and assistant routes. |
be-fast-and-stableefficient-resource-delivery | pass | high | Trace, HAR, layout and Lighthouse evidence directly measured loading, stability, main-thread and delivery behaviour across the landing and assistant routes. |
be-fast-and-stabletrim-unused-and-duplicate-code | issues | high | Its HAR transfers 583KB of script. Lighthouse estimated 235KiB unused JavaScript, 81KiB unminified JavaScript and 35KiB unused CSS. F4 medium: The assistant selector ships avoidable JavaScript and CSS. |
be-inclusivenames-roles-labels | pass | high | Lighthouse accessibility scored 1.00 on both the organisation and assistant routes; mobile screenshots and the chat DOM corroborate labelled controls, landmarks, live messaging and readable reflow. |
be-inclusivesufficient-contrast | pass | high | Lighthouse accessibility scored 1.00 on both the organisation and assistant routes; mobile screenshots and the chat DOM corroborate labelled controls, landmarks, live messaging and readable reflow. |
be-inclusivestructure-and-focus | pass | high | Lighthouse accessibility scored 1.00 on both the organisation and assistant routes; mobile screenshots and the chat DOM corroborate labelled controls, landmarks, live messaging and readable reflow. |
be-inclusivelegible-text | pass | high | Lighthouse accessibility scored 1.00 on both the organisation and assistant routes; mobile screenshots and the chat DOM corroborate labelled controls, landmarks, live messaging and readable reflow. |
be-inclusivezoom-reflow-targets-and-media | pass | high | Lighthouse accessibility scored 1.00 on both the organisation and assistant routes; mobile screenshots and the chat DOM corroborate labelled controls, landmarks, live messaging and readable reflow. |
follow-best-practicesno-console-errors | pass | high | Lighthouse, DOM and evaluate evidence confirms valid HTML/UTF-8, viewport, sized images and no landing-page console/best-practice failure; assistant platform exceptions are recorded. |
follow-best-practicessound-document-and-assets | pass | high | Lighthouse, DOM and evaluate evidence confirms valid HTML/UTF-8, viewport, sized images and no landing-page console/best-practice failure; assistant platform exceptions are recorded. |
follow-best-practicesbrowser-platform-hygiene | issues | high | Lighthouse reported one deprecated API, DevTools Issues panel entries, two BFCache failure reasons, and a best-practices score of 0.58. F5 medium: The assistant route triggers browser-platform hygiene failures. |
be-discoverabletitle-and-description | pass | high | Raw-HTML comparison found 100% content coverage on the public landing, sign-in guide and Self Assessment guide, with title, h1 and descriptions in raw HTML; canonical/JSON-LD/Open Graph and real hrefs were present. |
be-discoverablecrawlable-and-mobile-friendly | pass | high | Raw-HTML comparison found 100% content coverage on the public landing, sign-in guide and Self Assessment guide, with title, h1 and descriptions in raw HTML; canonical/JSON-LD/Open Graph and real hrefs were present. |
be-discoverablecanonical-and-indexing-signals | pass | high | Raw-HTML comparison found 100% content coverage on the public landing, sign-in guide and Self Assessment guide, with title, h1 and descriptions in raw HTML; canonical/JSON-LD/Open Graph and real hrefs were present. |
be-discoverablestructured-and-shareable-metadata | pass | high | Raw-HTML comparison found 100% content coverage on the public landing, sign-in guide and Self Assessment guide, with title, h1 and descriptions in raw HTML; canonical/JSON-LD/Open Graph and real hrefs were present. |
be-private-and-securesecure-transport-and-headers | issues | high | The cookie primitive found optimizelyOptOut scoped to .tax.service.gov.uk, Secure=false, SameSite=None and a 400-day expiry. HSTS mitigates many downgrade paths but does not justify an insecure cookie attribute. F6 medium: The assistant sets a long-lived cookie without the Secure flag. |
be-private-and-securedata-minimisation-and-third-parties | pass | high | HTTPS, nonce CSP, HSTS, nosniff and frame protection were observed; no exposed secrets or known tracker origins were found. The assistant-specific gaps are separately recorded. |
be-private-and-securein-context-permissions-and-modern-auth | pass | high | HTTPS, nonce CSP, HSTS, nosniff and frame protection were observed; no exposed secrets or known tracker origins were found. The assistant-specific gaps are separately recorded. |
be-private-and-securedefensive-browser-policies | issues | high | Headers evidence shows a CSP containing unsafe-inline and no Permissions-Policy on the assistant, while HSTS, nosniff and clickjacking protection are present. F7 medium: The assistant’s browser-enforced policy set is weaker than the GOV.UK landing page’s. |
be-resilientprogressive-enhancement | pass | high | Raw HTML preserved core public content; the tested 404 gives a clear “Page not found” state with recovery links, and the assistant selector/chat loaded without a JS shell. |
be-resilientresilient-runtime-behaviour | pass | high | Raw HTML preserved core public content; the tested 404 gives a clear “Page not found” state with recovery links, and the assistant selector/chat loaded without a JS shell. |
be-resilientoffline-and-installable | not-applicable | high | The public content is informational and the tax/chat task intrinsically requires a live government service; installability is not an expected core outcome. |
be-resilientnetwork-and-http-failure-states | pass | high | Raw HTML preserved core public content; the tested 404 gives a clear “Page not found” state with recovery links, and the assistant selector/chat loaded without a JS shell. |
be-internationalisedlang-dir-and-logical-properties | pass | high | HTML lang=en is correct and visible Welsh alternatives are offered; mobile reflow remains intact. |
be-internationalisedlocale-aware-data | not-applicable | high | The sampled pages do not render locale-sensitive dates, numbers, currencies, durations, or calendar data. |
be-internationalisedtime-zone-correctness | not-applicable | high | The sampled pages contain no event or time-zone workflow. |
be-trustworthyno-dark-patterns | pass | high | Cookie choices give equally prominent accept/reject controls; the chat warns against personal data, exposes End chat, labels input, and provides live-region announcements. |
be-trustworthyhumane-error-handling | pass | high | Cookie choices give equally prominent accept/reject controls; the chat warns against personal data, exposes End chat, labels input, and provides live-region announcements. |
be-trustworthytrustworthy-input-assistance | pass | high | Cookie choices give equally prominent accept/reject controls; the chat warns against personal data, exposes End chat, labels input, and provides live-region announcements. |
be-trustworthysafe-commercial-and-account-flows | not-applicable | high | No checkout, subscription, payment, cancellation, or sensitive account mutation was in the unauthenticated representative scope. |
be-sustainableoptimised-assets | issues | high | HAR measured 2.08MB of images out of 2.25MB total. The images primitive found 21/21 legacy-format images, no srcset, all below-fold images eager, and all natural dimensions over twice display width. F3 medium: The image-heavy organisation page transfers disproportionate legacy, non-responsive imagery. |
be-sustainableno-wasteful-work | pass | high | HAR, image and Lighthouse evidence measured page weight, third-party origins and background/main-thread work; issue rows capture the observed image waste. |
be-sustainablethird-party-and-media-budget | pass | high | HAR, image and Lighthouse evidence measured page weight, third-party origins and background/main-thread work; issue rows capture the observed image waste. |
be-agent-readystructured-agent-capabilities | not-applicable | high | No declared agent-facing capability is expected for this public information and support journey; this emerging check is contextual. |
be-agent-readyon-device-inference | not-applicable | high | No user task here calls for on-device inference; this emerging capability is contextual. |
be-memory-efficientno-leak-under-repeated-interaction | pass | medium | Chat heap snapshots were about 14.9MB baseline and 15.3MB after ten empty-send attempts; the bounded one-page footprint and constructor summaries showed no evidence of unbounded or detached-DOM growth in this short run. |
be-memory-efficientbounded-footprint | pass | medium | Chat heap snapshots were about 14.9MB baseline and 15.3MB after ten empty-send attempts; the bounded one-page footprint and constructor summaries showed no evidence of unbounded or detached-DOM growth in this short run. |
be-memory-efficientno-detached-dom-or-unbounded-listeners | pass | medium | Chat heap snapshots were about 14.9MB baseline and 15.3MB after ten empty-send attempts; the bounded one-page footprint and constructor summaries showed no evidence of unbounded or detached-DOM growth in this short run. |
Provenance
Canonical report: results/atomic/reports/0699-www_tax_service_gov_uk.json
Report SHA-256: d458313854ae793a57bc305532d1685b2564446cc23ef24e00beaff306098720
Local retained report: runs/2026-07-17T17-27-24-856Z/atomic-reports/www_tax_service_gov_uk/2026-07-22T11-42-31-501Z/report.json
Local evidence root: runs/2026-07-17T17-27-24-856Z/atomic-reports/www_tax_service_gov_uk/2026-07-22T11-42-31-501Z
Catalog SHA-256: sha256:78ccfdb2d483f4c57d9dafed80fd86c6265585a56457c8dcfddc254b80fb44d7
Raw screenshots, HARs, traces, heaps, and other browser artifacts are retained at the local evidence root and intentionally are not committed. Artifact paths in the canonical report are relative to that root.