Manifest position 90 · CrUX rank bucket 1000
https://www.amazon.es
Coverage complete
Atomic coverage complete across 58 checks and six representative paths. Anonymous audit; payment completion and signed-in account management were intentionally not covered.
All 58 atomic check outcomes
| Principle / check | Status | Confidence | Evidence or reason |
|---|---|---|---|
respect-user-preferencesrespects-color-scheme | issues | high | The dark-mode screenshot and computed-style probe show a white rgb(255,255,255) body, color-scheme: normal, and unchanged light surfaces while prefers-color-scheme: dark matches. F01 medium: The storefront ignores the user’s dark color-scheme preference. |
respect-user-preferencesrespects-reduced-motion | issues | high | The reduced-motion probe reports prefers-reduced-motion: reduce and three running 2,000 ms animations, the same count and duration as the ordinary preference probe. F02 medium: Non-essential homepage animation continues under reduced motion. |
respect-user-preferencesrespects-contrast | pass | high | Lighthouse contrast audit passes and the preference probe confirms prefers-contrast: more is active while text, controls, and focus remain visible. |
implement-natural-interactionsview-transitions | issues | medium | DOM/CSS probes on homepage, search, product, and basket find no view-transition CSS; the site is a multi-page commerce experience with frequent route changes. F03 low: Navigation changes are abrupt despite many route and product-detail transitions. |
implement-natural-interactionsscroll-driven-animations | not-applicable | high | No scroll-linked parallax, scrollytelling, or reveal interaction was present on the representative templates, so implementation technique is not applicable. |
implement-natural-interactionsphysical-gestures | pass | high | Representative shopping controls use native links, buttons, forms, and scrolling; no custom pointermove gesture that fights platform scrolling was observed. |
provide-guided-navigationscroll-state-aware-chrome | issues | medium | Full-page screenshots exceed 10,000 px in height, while CSS probes find no scroll-driven or scroll-state treatment and the header does not provide a visible progress or position response. F04 low: Long listing and product pages provide no scroll-state-aware navigation chrome. |
provide-guided-navigationanchored-positioning | issues | medium | The platform probe finds no anchor-name, position-anchor, or position-try CSS despite account/navigation menus and overlay controls. F05 low: Menus and overlays do not use resilient CSS anchor positioning. |
provide-guided-navigationdirects-attention | pass | high | Search/product templates provide breadcrumbs, skip links, descriptive section headings, and clear result/product destinations that preserve orientation. |
maximize-content-reduce-noiseno-intrusive-interruptions | issues | high | Desktop and mobile screenshots show a large fixed consent surface covering the center of the viewport; DOM identifies #cos-banner as a fixed form role=dialog. F06 high: The consent banner obscures primary shopping content on every first visit. |
maximize-content-reduce-noisesemantic-dismissible-primitives | issues | high | DOM reports zero <dialog> and popover elements; the fixed #cos-banner is a FORM with role=dialog and custom dismissal inputs. F07 medium: The consent modal is an ad-hoc form rather than a native dialog. |
maximize-content-reduce-noisereduced-chrome | issues | medium | The homepage full-page screenshot shows dozens of equally weighted card grids plus two navigation bars, while search renders 917 links and a very long filter/product grid. F08 medium: Dense repeated cards and commercial chrome overwhelm the content hierarchy. |
adapt-to-the-form-factorresponsive-no-horizontal-scroll | issues | high | At viewport 360x800 the layout primitive reports no viewport meta, innerWidth 1000, scroll content width 1000, and 640 px horizontal overflow; the mobile screenshot is a scaled desktop page. F09 critical: The desktop layout is forced into a 360 px mobile viewport. |
adapt-to-the-form-factorcomponent-level-responsiveness | issues | high | Across homepage/listing templates the probe finds zero container queries; mobile rendering scales the desktop grid instead of rearranging components for their available container. F10 medium: Reusable cards do not adapt at component level. |
adapt-to-the-form-factorinput-modality-aware | issues | high | Lighthouse target-size fails on footer links at 15 px high with insufficient spacing; the objective probe also lists numerous 14–20 px-high links. Focus outlines are visible, so the issue is specifically target sizing. F11 medium: Several interactive targets are too small for touch. |
support-core-task-successclear-purpose-and-primary-action | pass | high | Homepage search/navigation, search result heading, product title/price/add-to-basket, and basket state make purpose and next action clear. |
support-core-task-successprimary-flow-completion | pass | high | Search results expose real product hrefs and POST add-to-cart forms; product exposes Add to basket and basket exposes checkout entry. An automated add attempt navigated, confirming the form action, though post-navigation capture was not retained. |
support-core-task-successclear-system-state-and-recovery | pass | high | Empty basket clearly says it is empty and offers shopping recovery; the unknown route explains the problem and links home. |
be-fast-and-stablegood-core-web-vitals | issues | high | Lighthouse performance is 0.57 with FCP 7.5 s, LCP 14.3 s, and TTI 14.3 s. The unthrottled trace is much faster (LCP 1.69 s), so the failure is network/CPU-condition sensitive rather than universal. F12 high: Mobile lab loading performance is poor. |
be-fast-and-stablevisual-stability | pass | high | Layout observer and Lighthouse both report CLS 0 with no captured shift entries. |
be-fast-and-stableefficient-main-thread | pass | high | Trace records 3 long tasks, 118 ms TBT on homepage; Lighthouse TBT is 130 ms, within a good lab range. |
be-fast-and-stableefficient-resource-delivery | issues | high | Homepage HAR transfers 3.33 MB in 167 requests, including 1.02 MB script, 722 KB fonts, and multiple VeryHigh-priority parser-inserted stylesheets confirmed in the document head. F13 medium: The critical network path is heavy and stylesheet-dominated. |
be-fast-and-stabletrim-unused-and-duplicate-code | issues | high | Lighthouse estimates 524 KiB unused JavaScript and 185 KiB unused CSS, including a 590 KB WAF challenge script with about 328 KB unused. F14 medium: The homepage ships substantial unused JavaScript and CSS. |
be-inclusivenames-roles-labels | issues | high | Lighthouse accessibility finds aria-expanded invalid on the role=searchbox input, aria-setsize invalid on a list, and the visible cart count missing from the cart link accessible name. F15 medium: ARIA roles and accessible names are inconsistent. |
be-inclusivesufficient-contrast | pass | high | Lighthouse color-contrast audit passes with no failing nodes. |
be-inclusivestructure-and-focus | issues | high | Homepage has no H1, search has no main landmark, and the product page exposes five H1 elements. Focus is visible, but the document outline does not consistently identify one primary topic and main region. F16 medium: Heading and landmark structure varies unpredictably by template. |
be-inclusivelegible-text | pass | medium | Desktop screenshots show consistently rendered, unclipped text with clear headings and labels across sampled templates. |
be-inclusivezoom-reflow-targets-and-media | issues | high | The 360 px layout has 640 px horizontal overflow and no viewport meta; Lighthouse independently fails target-size. This prevents a usable zoom/reflow experience. F17 critical: The page fails mobile reflow and target-size requirements. |
follow-best-practicesno-console-errors | pass | high | Lighthouse reports no browser console errors. |
follow-best-practicessound-document-and-assets | issues | high | The image primitive finds 42 of 43 images missing dimensions; Lighthouse identifies six incorrectly displayed aspect ratios. F18 medium: Image sizing and aspect-ratio hygiene is weak. |
follow-best-practicesbrowser-platform-hygiene | issues | high | Lighthouse reports that the Shared Storage API used from the main document is deprecated and will be removed. F19 low: The page uses a deprecated Shared Storage API surface. |
be-discoverabletitle-and-description | issues | high | Homepage and product have descriptions, but objective probes return description:null for search results and the basket. F20 medium: Search and basket templates omit meta descriptions. |
be-discoverablecrawlable-and-mobile-friendly | issues | high | No representative page has a viewport meta tag, and Lighthouse finds a javascript:void(0) anchor that is not crawlable. F21 high: Mobile and crawlability signals are incomplete. |
be-discoverablecanonical-and-indexing-signals | issues | high | Homepage/product expose canonicals, but search has none, /sitemap.xml returns 404, and no hreflang links are exposed despite localized language routes. robots.txt allows general indexing but separately blocks many AI user agents. F22 medium: Indexing signals are inconsistent across templates. |
be-discoverablestructured-and-shareable-metadata | issues | high | The product probe finds zero JSON-LD blocks and no Open Graph or Twitter metadata despite a visible product entity with price, reviews, availability, and images. F23 high: Product pages lack machine-readable product and share metadata. |
be-private-and-securesecure-transport-and-headers | issues | high | Headers primitive finds no CSP, HSTS, nosniff, frame protection, Referrer-Policy, or Permissions-Policy. Cookies primitive reports seven of eight cookies with security concerns, including non-Secure cookies and session-like cookies without HttpOnly. F24 high: Transport is HTTPS, but baseline security headers and cookie flags are missing. |
be-private-and-securedata-minimisation-and-third-parties | issues | high | The consent notice names 129 third parties; the HAR records 147 cross-origin requests and 2.70 MB cross-origin transfer, including ads/telemetry endpoints, before the banner is acted on. Many media origins are Amazon-owned, but the pre-choice request volume remains substantial. F25 high: The first-load data and third-party footprint is large before a consent choice. |
be-private-and-securein-context-permissions-and-modern-auth | pass | high | No permission prompt appears on load; sign-in uses username autocomplete and includes passkey recovery/error states in its DOM alongside password fallback. |
be-private-and-securedefensive-browser-policies | issues | high | The headers capture has no CSP/frame-ancestors or X-Frame-Options, HSTS, nosniff, Referrer-Policy, or Permissions-Policy. F26 high: Browser-enforced defensive policies are absent. |
be-resilientprogressive-enhancement | issues | high | Discoverability fetch receives HTTP 202 with only 2,007 bytes and 1% rendered-word coverage; title and description are absent and the crawler view is a shell. Product detail is better at 47%, showing the failure is template/edge-policy dependent. F27 high: The homepage is effectively unavailable to a non-JavaScript crawler. |
be-resilientresilient-runtime-behaviour | pass | high | Menus, consent, listing, product, basket, and error templates render complete, usable states without clipped primary controls at desktop. |
be-resilientoffline-and-installable | not-applicable | high | The public storefront is commerce rather than an installable web app; offline checkout is intrinsically network-dependent, so installability/offline fallback is not required. |
be-resilientnetwork-and-http-failure-states | pass | medium | The sampled missing route returns a clear localized not-found state and a link back to the Amazon homepage; basket empty state also offers recovery. |
be-internationalisedlang-dir-and-logical-properties | pass | high | Sampled pages declare lang=en-gb and resolve ltr direction consistently while localized /-/en/ links preserve language context. |
be-internationalisedlocale-aware-data | pass | high | The English locale renders euro pricing and localized commercial text consistently across product/search paths. |
be-internationalisedtime-zone-correctness | not-applicable | high | No user-visible scheduled event or time-zone-sensitive value appeared on the representative anonymous commerce paths. |
be-trustworthyno-dark-patterns | pass | high | Consent presents adjacent Accept, Decline, and Customise controls without confirmshaming; product pricing and purchase actions are visibly labelled. |
be-trustworthyhumane-error-handling | pass | high | Sign-in includes specific invalid mobile/email, generic retry, and passkey recovery messages; 404 and empty basket states explain recovery. |
be-trustworthytrustworthy-input-assistance | pass | high | The active sign-in identity field is type=email with autocomplete=username and an associated label; search fields are labelled. |
be-trustworthysafe-commercial-and-account-flows | pass | medium | Product price, quantity/purchase choices, Add to basket, Buy Now, empty basket, and sign-in commitment text are explicit on sampled anonymous paths. |
be-sustainableoptimised-assets | issues | high | The image primitive finds 40 images without srcset, 33 legacy-format images, zero modern formats, 24 below-fold images not lazy-loaded, and 42 missing dimensions. F28 medium: Image delivery omits basic responsive and modern-format optimisations. |
be-sustainableno-wasteful-work | issues | high | Homepage makes 167 requests and search makes 137; telemetry endpoints continue alongside a 25-script homepage payload and Lighthouse identifies large unused code. F29 medium: The storefront performs more background/network work than the initial task needs. |
be-sustainablethird-party-and-media-budget | issues | high | Homepage transfers 3.33 MB: 1.02 MB script, 839 KB images, and 722 KB across 21 font requests; 2.70 MB is cross-origin by hostname. F30 medium: Fonts, scripts, and cross-origin assets consume a disproportionate first-load budget. |
be-agent-readystructured-agent-capabilities | not-applicable | high | WebMCP is emerging and Amazon explicitly blocks the sampled AI crawler user agents in robots.txt; no declared agent-facing surface was inferred. |
be-agent-readyon-device-inference | not-applicable | high | No user task in the sampled shopping paths requires on-device inference; absence is an opportunity, not a failure. |
be-memory-efficientno-leak-under-repeated-interaction | pass | medium | After 10 account-menu hover cycles, heap summary is slightly lower than baseline (32,471,736 vs 32,574,106 bytes; 453,176 vs 456,814 nodes). No retained-growth signal appears in this bounded sample. |
be-memory-efficientbounded-footprint | pass | medium | Heap summary is about 32.6 MB self size for a complex commerce page; Performance memory is about 21–23 MB used and DOM count about 2,003, proportionate to the sampled page. |
be-memory-efficientno-detached-dom-or-unbounded-listeners | pass | high | Before/post summaries show no growing Detached* constructor among top constructors, node count decreases, and repeated menu interaction adds only one live DOM node. |
Provenance
Canonical report: results/atomic/reports/0090-www_amazon_es.json
Report SHA-256: 03e2a9b595a387cb987970ddbba926b381c38cc307d1f492e4b523d4826d04df
Local retained report: runs/2026-07-17T17-27-24-856Z/atomic-reports/www_amazon_es/2026-07-18T08-22-24-388Z/report.json
Local evidence root: runs/2026-07-17T17-27-24-856Z/atomic-reports/www_amazon_es/2026-07-18T08-22-24-388Z
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.