Manifest position 207 · CrUX rank bucket 1000
https://zh.wikipedia.org
Coverage complete
Coverage-complete audit of five representative public templates/flows. Editing, authenticated contribution, donation/payment, and media-playback flows were not covered because they require state, credentials, or cross-site commitment beyond this public audit.
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 screenshot captured with prefers-color-scheme: dark remains a white, light-themed page, despite the CSS feature probe finding color-scheme rules. F001 medium: The default experience does not follow the system dark preference. |
respect-user-preferencesrespects-reduced-motion | pass | medium | evaluate under prefers-reduced-motion: reduce found zero active animations and zero nonzero control transitions |
respect-user-preferencesrespects-contrast | pass | medium | high-contrast screenshot remained readable with visible controls; CSS probe found prefers-contrast/forced-colors support |
implement-natural-interactionsview-transitions | issues | medium | The stylesheet probe found no view-transition usage across the representative article while navigation is a central repeated interaction. F015 low: Article and route changes use abrupt full-document swaps without View Transitions. |
implement-natural-interactionsscroll-driven-animations | pass | medium | stylesheet probe found no scroll-linked animation and the site uses native document scrolling, avoiding main-thread scroll animation work |
implement-natural-interactionsphysical-gestures | pass | medium | screenshots and DOM show standard browser scrolling and controls with no custom swipe/pointer gesture surface on tested paths |
provide-guided-navigationscroll-state-aware-chrome | issues | medium | The long article screenshot and stylesheet probe show persistent table-of-contents/chrome but no scroll-state container query or scroll progress treatment. F016 low: Long-form navigation chrome does not visibly adapt to scroll state. |
provide-guided-navigationanchored-positioning | pass | medium | tested overlays are centered preference UI or fixed menus with no anchor-dependent tooltip placement requirement observed |
provide-guided-navigationdirects-attention | pass | medium | article has a visible table of contents, headings, selected page tab, skip link, and clear link styling |
maximize-content-reduce-noiseno-intrusive-interruptions | issues | high | Desktop and mobile screenshots show a large prompt over the homepage/article before interaction. On mobile it occupies the bottom of an already scaled-down viewport. F002 medium: A language-variant prompt obscures core content on first load. |
maximize-content-reduce-noisesemantic-dismissible-primitives | issues | high | The prompt is visible in screenshots, while the article DOM probe reports 0 dialog elements and 0 popover elements. F003 medium: The visible modal-like preference prompt is not exposed with a native dialog or popover primitive. |
maximize-content-reduce-noisereduced-chrome | pass | medium | article content remains the dominant central surface and optional side chrome can be hidden |
adapt-to-the-form-factorresponsive-no-horizontal-scroll | issues | high | The DOM reports meta viewport width=1120. The 360x800 layout records clientWidth 1120 and visual scale 0.321, and mobile screenshots show tiny desktop columns rather than reflowed content. F004 critical: The site forces a 1120 CSS-pixel desktop layout into a 360-pixel mobile viewport. |
adapt-to-the-form-factorcomponent-level-responsiveness | issues | high | The DOM reports meta viewport width=1120. The 360x800 layout records clientWidth 1120 and visual scale 0.321, and mobile screenshots show tiny desktop columns rather than reflowed content. F004 critical: The site forces a 1120 CSS-pixel desktop layout into a 360-pixel mobile viewport. |
adapt-to-the-form-factorinput-modality-aware | issues | high | Lighthouse reports target-size failures. The focus probe found multiple visible buttons, links, and the search input with outline none and no compensating box shadow. F005 high: Touch targets and focus presentation are inconsistent. |
support-core-task-successclear-purpose-and-primary-action | pass | medium | homepage, article, search and login screenshots each expose a descriptive heading and clear next action |
support-core-task-successprimary-flow-completion | pass | medium | article reading, exact search redirect, empty search, and login entry were reachable without dead ends |
support-core-task-successclear-system-state-and-recovery | pass | medium | empty search and missing-article paths explain the state and offer search, edit, refresh, and onward links |
be-fast-and-stablegood-core-web-vitals | issues | high | Lighthouse performance is 0.58 with FCP 4.2s, LCP 8.8s, Speed Index 5.3s, and TBT 210ms. The separate desktop trace was healthier at LCP 1.68s and TBT 24.81ms, showing a form-factor/network sensitivity. F006 high: The representative article loads too slowly under Lighthouse mobile conditions. |
be-fast-and-stablevisual-stability | pass | medium | layout observer measured CLS 0.0199 and Lighthouse measured 0.0849, both within the good threshold |
be-fast-and-stableefficient-main-thread | issues | high | Lighthouse performance is 0.58 with FCP 4.2s, LCP 8.8s, Speed Index 5.3s, and TBT 210ms. The separate desktop trace was healthier at LCP 1.68s and TBT 24.81ms, showing a form-factor/network sensitivity. F006 high: The representative article loads too slowly under Lighthouse mobile conditions. |
be-fast-and-stableefficient-resource-delivery | issues | high | HAR captured 54 requests and 1,055,413 transferred bytes. Scripts account for 635,308 bytes, including a 357,929-byte bundle, and three parser-inserted VeryHigh-priority stylesheets are render-blocking candidates. F007 medium: Critical delivery includes large scripts and render-blocking styles. |
be-fast-and-stabletrim-unused-and-duplicate-code | issues | high | Lighthouse estimates 165 KiB unused JavaScript and 94 KiB unused CSS on the representative article. F008 medium: The article ships substantial unused JavaScript and CSS. |
be-inclusivenames-roles-labels | issues | high | Lighthouse fails image-alt, label, and link-name audits. The image probe reports 74 of 85 img elements without alt, including many content images, though some may be decorative. F009 high: Some images, controls, and links lack accessible alternatives or names. |
be-inclusivesufficient-contrast | pass | medium | Lighthouse accessibility review did not fail color-contrast and the high-contrast screenshot remains legible |
be-inclusivestructure-and-focus | issues | high | The focus probe found several buttons, links, and inputs with computed outline none and no visible box-shadow. Lighthouse also reports unnamed links, weakening navigation context. F010 medium: Keyboard focus is not consistently visible across interactive chrome. |
be-inclusivelegible-text | issues | high | The DOM reports meta viewport width=1120. The 360x800 layout records clientWidth 1120 and visual scale 0.321, and mobile screenshots show tiny desktop columns rather than reflowed content. F004 critical: The site forces a 1120 CSS-pixel desktop layout into a 360-pixel mobile viewport. |
be-inclusivezoom-reflow-targets-and-media | issues | high | The DOM reports meta viewport width=1120. The 360x800 layout records clientWidth 1120 and visual scale 0.321, and mobile screenshots show tiny desktop columns rather than reflowed content. Lighthouse reports target-size failures. The focus probe found multiple visible buttons, links, and the search input with outline none and no compensating box shadow. F004 critical: The site forces a 1120 CSS-pixel desktop layout into a 360-pixel mobile viewport. F005 high: Touch targets and focus presentation are inconsistent. |
follow-best-practicesno-console-errors | pass | medium | Lighthouse errors-in-console audit scored 1 with an empty error table |
follow-best-practicessound-document-and-assets | issues | high | The image probe reports 73 below-fold images without loading=lazy, 79 legacy-format images, 3 missing dimensions, 5 missing srcset, and 2 oversized assets. F013 medium: Article media is not consistently lazy-loaded, modern-formatted, or fully dimensioned. |
follow-best-practicesbrowser-platform-hygiene | issues | high | Lighthouse reports missing source maps for large first-party JavaScript and Chrome inspector issues, lowering best-practices to 0.77. F014 low: Developer-facing browser hygiene is incomplete. |
be-discoverabletitle-and-description | issues | high | The article DOM probe reports description null and Lighthouse fails meta-description, although the title is descriptive. F011 medium: Public article pages omit a meta description. |
be-discoverablecrawlable-and-mobile-friendly | issues | high | The DOM reports meta viewport width=1120. The 360x800 layout records clientWidth 1120 and visual scale 0.321, and mobile screenshots show tiny desktop columns rather than reflowed content. F004 critical: The site forces a 1120 CSS-pixel desktop layout into a 360-pixel mobile viewport. |
be-discoverablecanonical-and-indexing-signals | pass | medium | status 200 content has canonical and 10 hreflang links; robots.txt and sitemap.xml were fetched successfully |
be-discoverablestructured-and-shareable-metadata | pass | medium | article probe found one JSON-LD block and five Open Graph properties matching visible article content |
be-private-and-securesecure-transport-and-headers | issues | high | Headers evidence finds CSP with unsafe-inline and unsafe-eval, no frame-ancestors/X-Frame-Options, no response Referrer-Policy, and no Permissions-Policy. Cookie evidence finds a non-Secure zhwikimwuser-sessionId cookie and six SameSite=None cookies. F012 high: Browser defenses are incomplete and the CSP remains broadly permissive. |
be-private-and-securedata-minimisation-and-third-parties | pass | medium | tracker capture found no known tracker domains; third parties were Wikimedia infrastructure only |
be-private-and-securein-context-permissions-and-modern-auth | pass | medium | no permission prompt appeared on load; login uses username webauthn and current-password autocomplete tokens |
be-private-and-securedefensive-browser-policies | issues | high | Headers evidence finds CSP with unsafe-inline and unsafe-eval, no frame-ancestors/X-Frame-Options, no response Referrer-Policy, and no Permissions-Policy. Cookie evidence finds a non-Secure zhwikimwuser-sessionId cookie and six SameSite=None cookies. F012 high: Browser defenses are incomplete and the CSP remains broadly permissive. |
be-resilientprogressive-enhancement | pass | medium | discoverability capture found 100% rendered-content coverage in raw HTML and a fully meaningful crawler view |
be-resilientresilient-runtime-behaviour | pass | medium | navigation, preference chrome, search, and recovery states rendered without cut-off desktop overlays or blank runtime states |
be-resilientoffline-and-installable | not-applicable | high | Wikipedia is a public content site rather than an installable app; offline installation was judged outside this representative scope. |
be-resilientnetwork-and-http-failure-states | pass | medium | missing content and empty search both provide explicit, actionable recovery rather than a blank page |
be-internationalisedlang-dir-and-logical-properties | pass | medium | html lang=zh, dir=ltr, 10 hreflang variants, and logical CSS properties were observed |
be-internationalisedlocale-aware-data | pass | low | Chinese-language article, date, count, and locale variants render in localized notation across tested pages |
be-internationalisedtime-zone-correctness | not-applicable | high | The audited reading, search, and login paths expose no scheduled event or user-entered time workflow to test. |
be-trustworthyno-dark-patterns | pass | medium | no ads, forced continuity, confirmshaming, checkout, or coercive consent appeared on any tested path |
be-trustworthyhumane-error-handling | pass | medium | empty-search and missing-content errors use neutral language and give actionable recovery choices |
be-trustworthytrustworthy-input-assistance | pass | medium | login username uses username webauthn and password uses current-password autocomplete with visible labels |
be-trustworthysafe-commercial-and-account-flows | pass | medium | login is clearly labelled, offers password recovery and account creation, and does not expose commercial commitments |
be-sustainableoptimised-assets | issues | high | The image probe reports 73 below-fold images without loading=lazy, 79 legacy-format images, 3 missing dimensions, 5 missing srcset, and 2 oversized assets. F013 medium: Article media is not consistently lazy-loaded, modern-formatted, or fully dimensioned. |
be-sustainableno-wasteful-work | issues | high | Lighthouse estimates 165 KiB unused JavaScript and 94 KiB unused CSS on the representative article. F008 medium: The article ships substantial unused JavaScript and CSS. |
be-sustainablethird-party-and-media-budget | pass | medium | HAR found no video/audio/autoplay and third-party bytes were Wikimedia-hosted content infrastructure; no known trackers were found |
be-agent-readystructured-agent-capabilities | not-applicable | high | The emerging WebMCP-style capability check was judged not applicable to these public read/search pages; existing crawler and API access were assessed under discoverability. |
be-agent-readyon-device-inference | not-applicable | high | No user task in the audited reading, search, or login surfaces requires on-device inference. |
be-memory-efficientno-leak-under-repeated-interaction | pass | low | after 10 repeated menu toggles the heap remained bounded at 23.1 MB with no Detached constructor among top populations; comparison is low-confidence because snapshots used isolated sessions |
be-memory-efficientbounded-footprint | pass | medium | heap summaries measured 20.2 to 23.1 MB self size for a media-rich portal, proportionate to the tested page |
be-memory-efficientno-detached-dom-or-unbounded-listeners | pass | medium | baseline and post summaries exposed no Detached constructor among top populations after repeated menu toggles |
Provenance
Canonical report: results/atomic/reports/0207-zh_wikipedia_org.json
Report SHA-256: 212bf50b59f7ca5bb7c66055be7e7abd5bc4fbcb554773a40f43fb72b82195b5
Local retained report: runs/2026-07-17T17-27-24-856Z/atomic-reports/zh_wikipedia_org/2026-07-19T04-19-20-711Z/report.json
Local evidence root: runs/2026-07-17T17-27-24-856Z/atomic-reports/zh_wikipedia_org/2026-07-19T04-19-20-711Z
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.