Manifest position 679 · CrUX rank bucket 1000
https://search.naver.com
Coverage complete
Atomic coverage complete across four representative surfaces. The bare host redirects to www.naver.com; that homepage and three search templates were included.
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 | Dark and default screenshots are byte-identical; all principal surfaces stay white under prefers-color-scheme: dark. F001 medium: No system-driven dark theme |
respect-user-preferencesrespects-reduced-motion | issues | high | The default and reduce probes both report the same 300ms animation and 50 transitioning elements; loading-dot animations also remain defined. F002 medium: Reduced-motion preference does not reduce motion |
respect-user-preferencesrespects-contrast | issues | high | The prefers-contrast screenshot is visually unchanged while Lighthouse finds 23 low-contrast nodes, including tabs at 3.36:1. F003 medium: High-contrast preference is not reflected |
implement-natural-interactionsview-transitions | issues | high | CSS inspection found no view-transition rules; the autocomplete interaction uses ordinary transitions and state replacement. F004 low: State and route changes omit View Transitions |
implement-natural-interactionsscroll-driven-animations | not-applicable | high | No scroll-linked decorative/parallax/scrollytelling motion exists on the audited search surfaces, so there is no scroll animation implementation to assess. |
implement-natural-interactionsphysical-gestures | pass | high | Screenshots and DOM inspection show native scrolling, links, tabs, and buttons; no custom pointer-driven gesture surface that fights platform scrolling was found. |
provide-guided-navigationscroll-state-aware-chrome | issues | high | The results page is over 5,300 CSS px tall, but CSS inspection found no scroll-state query, progress indicator, or equivalent position-aware chrome. F005 low: Long results do not expose scroll-state guidance |
provide-guided-navigationanchored-positioning | issues | high | The DOM contains four custom role=dialog overlays, no popover elements, and no anchor-name/position-anchor/position-try CSS. F006 low: Overlays use custom positioning rather than anchored primitives |
provide-guided-navigationdirects-attention | pass | high | The results page has selected tab state, two skip links, descriptive section headings, and visible result-card grouping that preserves orientation. |
maximize-content-reduce-noiseno-intrusive-interruptions | issues | high | The initial homepage screenshot shows both a top install banner and a large floating Whale install panel obscuring the content before user interaction. F007 high: Homepage interrupts content with browser-install promotion |
maximize-content-reduce-noisesemantic-dismissible-primitives | issues | high | Results expose four role=dialog containers but zero dialog, popover, or details elements. F008 medium: Custom overlay semantics replace native primitives |
maximize-content-reduce-noisereduced-chrome | issues | high | The first viewport devotes substantial area to install promotions, service icons, a large ad, and news chrome around the search field. F009 medium: Homepage chrome and promotions crowd the core task |
adapt-to-the-form-factorresponsive-no-horizontal-scroll | issues | high | The desktop response has no viewport meta; layout measured 1,160px content in a 780px viewport with 380px overflow, and a 360px capture is scaled down rather than genuinely reflowed. F010 high: Desktop search response lacks a viewport contract and fixed-width layout overflows |
adapt-to-the-form-factorcomponent-level-responsiveness | issues | high | CSS inspection found no @container rules despite cards, AI panels, tabs, and side modules appearing in varied widths. F011 medium: Complex reused result components are not container-responsive |
adapt-to-the-form-factorinput-modality-aware | issues | high | The active probe counted 58 controls under 44px in one dimension; several focused controls report outline:none, and Lighthouse confirms undersized 20px targets. F012 high: Many controls are too small or lose visible focus |
support-core-task-successclear-purpose-and-primary-action | pass | high | The first viewport contains a prominent query field, search button, query-specific title, and clearly grouped results. |
support-core-task-successprimary-flow-completion | pass | high | A GET search query reaches a 200 result page end-to-end, with navigable result links and category tabs; the replayed input interaction remained usable. |
support-core-task-successclear-system-state-and-recovery | issues | high | A deliberately nonsensical query immediately shows product ads/comparisons without a clear no-results explanation, query correction, or primary recovery action in the first viewport. F013 medium: No-match search state presents unrelated commerce before clear recovery |
be-fast-and-stablegood-core-web-vitals | issues | high | Lighthouse measured FCP 6.3s, LCP 30.5s, TBT 1,340ms, and performance 0.29; the desktop trace also measured LCP 4.45s. F014 critical: Mobile search is far outside good Core Web Vitals |
be-fast-and-stablevisual-stability | issues | high | Layout evidence measured CLS 0.163 on desktop results and 0.769 on the redirected homepage mobile condition. F015 high: Late content causes substantial layout shift |
be-fast-and-stableefficient-main-thread | issues | high | The trace found four long tasks, longest 138ms, with 228ms TBT; mobile Lighthouse reported 1,340ms TBT. F016 high: Heavy scripting blocks the main thread |
be-fast-and-stableefficient-resource-delivery | issues | high | HAR captured 85 requests and 4.56MB transferred, including eight parser-inserted high-priority stylesheets and several high-priority scripts; trace LCP was 4.45s. F017 high: Critical delivery is heavy and render-blocking |
be-fast-and-stabletrim-unused-and-duplicate-code | issues | high | HAR shows two separate index.js bundles at about 1.08MB each; scripts total 3.63MB across 41 requests. F018 high: Two megabyte-scale application bundles dominate the load |
be-inclusivenames-roles-labels | issues | high | Lighthouse found 14 images without alt, four unnamed links, one unnamed button, and three elements with disallowed aria-expanded; image search found 24 missing alt. F019 high: Controls and images lack accessible names or valid semantics |
be-inclusivesufficient-contrast | issues | high | Lighthouse found 23 low-contrast nodes; search tabs use #8c8c8c on white at 3.36:1 instead of 4.5:1. F020 high: Text and tabs fail minimum contrast |
be-inclusivestructure-and-focus | issues | high | Mobile Lighthouse found no main landmark and invalid ARIA; the focus probe shows the search input and multiple custom controls with outline:none. F021 high: Landmarks, ARIA state, and focus treatment are inconsistent |
be-inclusivelegible-text | pass | high | Desktop and mobile result screenshots show unclipped result titles/snippets with clear grouping and adequate line spacing; no text clipping was observed. |
be-inclusivezoom-reflow-targets-and-media | issues | high | The desktop response omits viewport meta, scales a fixed-width page at 360px, and Lighthouse reports insufficient 20px touch targets. F022 high: Reflow and target sizing are not robust |
follow-best-practicesno-console-errors | pass | high | Lighthouse errors-in-console audit scored 1 and reported no console errors. |
follow-best-practicessound-document-and-assets | pass | high | Probe and Lighthouse confirm HTML doctype, UTF-8 charset, correct image aspect ratios, and explicit dimensions on all 75 image-search images. |
follow-best-practicesbrowser-platform-hygiene | issues | high | Lighthouse reports an unload event listener in nx-m-1.1.3.min.js, an API Chrome is removing. F023 medium: Deprecated unload handling risks BFCache and platform compatibility |
be-discoverabletitle-and-description | pass | high | Probe confirms a query-specific title and meta description; raw HTML retains both and discoverability coverage is 85% on results. |
be-discoverablecrawlable-and-mobile-friendly | issues | high | The desktop result omits viewport meta, Lighthouse flags four generic “자세히 보기” links, and the search host intentionally blocks indexing. F024 medium: Search response has weak mobile/crawlability details |
be-discoverablecanonical-and-indexing-signals | not-applicable | high | Search result URLs are intentionally excluded from indexing by robots.txt to avoid indexing arbitrary query combinations; canonical/hreflang indexing is not applicable to this internal-search template. |
be-discoverablestructured-and-shareable-metadata | not-applicable | high | Generic search result and homepage templates do not represent a single article, product, event, place, or other rich entity requiring schema.org entity markup. |
be-private-and-securesecure-transport-and-headers | issues | high | HTTPS and CSP are present, but HSTS and nosniff are absent; three first-party cookies lack Secure and all seven use SameSite=None. F025 high: Transport policy and cookies have avoidable gaps |
be-private-and-securedata-minimisation-and-third-parties | issues | high | Tracker evidence saw 14 cross-origin hosts; HAR attributes 81/85 requests and 4.48MB to other origins, including logging, impression, and responsive telemetry endpoints. F026 medium: The search load has a broad telemetry and cross-origin footprint |
be-private-and-securein-context-permissions-and-modern-auth | not-applicable | high | No permission request or authentication flow occurs on the audited anonymous search surfaces; only a link to a separate sign-in service is present. |
be-private-and-securedefensive-browser-policies | issues | high | Headers evidence confirms frame-ancestors and X-Frame-Options, but no HSTS, Permissions-Policy, or X-Content-Type-Options. F027 medium: Defensive browser headers are incomplete |
be-resilientprogressive-enhancement | issues | high | Discoverability measured only 6% raw/rendered word coverage on www.naver.com with an empty #root; the crawler screenshot is effectively blank, while results reach 85%. F028 high: The redirected homepage is an empty JavaScript shell to non-JS clients |
be-resilientresilient-runtime-behaviour | pass | high | The homepage promotion and search overlays remain within the viewport and expose close controls; repeated autocomplete focus/input cycles did not break the DOM state. |
be-resilientoffline-and-installable | not-applicable | high | A live web search intrinsically depends on current network results; installability/offline result delivery is not a reasonable core requirement, although useful network failure UI remains applicable. |
be-resilientnetwork-and-http-failure-states | pass | high | The invalid sitemap/error response presents a clear Korean error message plus previous-page and Naver-home recovery links instead of a blank shell. |
be-internationalisedlang-dir-and-logical-properties | pass | high | The rendered and raw documents declare lang=ko and UTF-8, matching the Korean content and correct left-to-right reading order. |
be-internationalisedlocale-aware-data | not-applicable | high | The audited Korean search surfaces show no user-facing dates, currencies, durations, or other locale-sensitive values requiring alternate formatting. |
be-internationalisedtime-zone-correctness | not-applicable | high | No events, schedules, stored times, recurring intervals, or time-zone-sensitive data appear in the representative search surfaces. |
be-trustworthyno-dark-patterns | pass | high | Across search, no-match, and image-result screenshots, advertising is visually distinguishable and no forced continuity, confirmshaming, or hidden cancellation flow was observed. |
be-trustworthyhumane-error-handling | not-applicable | high | The primary search form has no required structured fields or invalid-value state; arbitrary text is a valid query. |
be-trustworthytrustworthy-input-assistance | pass | high | The search input exposes a combobox role and the recorded interaction displays query assistance without blocking typing or submission. |
be-trustworthysafe-commercial-and-account-flows | not-applicable | high | Checkout, subscription, consent, authentication, and account-management flows are outside the anonymous search templates audited. |
be-sustainableoptimised-assets | issues | high | The images probe found 50 below-fold images without lazy loading and 48 images without srcset/responsive sources. F029 medium: Image result delivery misses responsive and lazy-loading opportunities |
be-sustainableno-wasteful-work | issues | high | Scripts transfer 3.63MB across 41 requests, including duplicate 1.08MB bundles, while several logging/impression calls run during initial load. F030 high: Large duplicate scripts and background telemetry waste device/network work |
be-sustainablethird-party-and-media-budget | issues | high | HAR reports 4.48MB of 4.56MB from other origins, with large audio and image-viewer scripts loaded on a general result page. F031 high: Cross-origin script and media cost is disproportionate for one result page |
be-agent-readystructured-agent-capabilities | not-applicable | high | The host explicitly disallows AI retrieval bots in robots.txt and exposes no declared agent-facing surface, so structured agent tools are judged intentionally out of scope. |
be-agent-readyon-device-inference | not-applicable | high | The host declares no agent-facing intent and the audited search task does not require a built-in on-device model; absence is an opportunity, not a failure. |
be-memory-efficientno-leak-under-repeated-interaction | pass | high | After ten representative autocomplete focus/input/clear cycles, heap self size fell from 90,823,662 to 90,739,743 bytes; FiberNode count stayed 13,265 and live DOM nodes stayed 3,049. |
be-memory-efficientbounded-footprint | issues | high | The baseline heap summary reports 90.8MB self size, 1.45M heap nodes, 118,512 closures, and 13,265 FiberNode objects for about 3,049 live DOM elements. F032 medium: Initial heap footprint is heavy for a search-results document |
be-memory-efficientno-detached-dom-or-unbounded-listeners | pass | high | Before/after heap summaries show no Detached* constructor among top populations; closures rose only 5 and DOM/animation counts remained unchanged after ten cycles. |
Provenance
Canonical report: results/atomic/reports/0679-search_naver_com.json
Report SHA-256: e439d77d8660991efcd6c5fb628b9489bcb1a6a2f296823170b43a1cf15991d2
Local retained report: runs/2026-07-17T17-27-24-856Z/atomic-reports/search_naver_com/2026-07-22T08-49-56-085Z/report.json
Local evidence root: runs/2026-07-17T17-27-24-856Z/atomic-reports/search_naver_com/2026-07-22T08-49-56-085Z
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.