{ "url": "https://es.scribd.com", "auditedAt": "2026-07-23T10:23:22.763Z", "mode": "report", "status": "completed", "statusDetail": "Coverage-complete model audit across five representative public templates/journeys. The requested Spanish host redirects most inspected paths to English; this is reported as a finding rather than an execution blocker.", "page": { "appType": "hybrid", "framework": "Next.js/React-like marketing and category surfaces plus Auth0-hosted signup", "notes": "Public document discovery/subscription service. Covered home, category, search, detail, and signup. Did not submit personal data, create an account, purchase, upload a document, or test authenticated reading/cancellation." }, "evidenceUsed": [ "screenshot", "dom", "evaluate-probe", "layout-metrics", "trace", "har", "discoverability", "images", "headers", "cookies", "trackers", "secrets", "lighthouse+axe-core", "heap-summary", "robots.txt", "modern-web-guidance" ], "guidanceConsulted": [ "accessibility", "privacy", "security", "forms", "dark-mode", "performance", "fluid-scaling", "optimize-image-priority", "break-up-long-tasks", "light-dismiss-a-dialog", "platform-controls-dismiss-dialog", "improve-text-layout-and-legibility", "validate-input-after-interaction", "required-field-feedback", "passkeys", "translator", "language-detection", "deprioritize-background-fetches", "efficient-background-processing", "webmcp", "html", "defer-rendering-heavy-content" ], "artifacts": [ { "type": "screenshot", "path": "evidence/category-desktop.png", "caption": "category desktop", "findingIds": [ "F2" ] }, { "type": "screenshot", "path": "evidence/category-discoverability-crawler.png", "caption": "category discoverability crawler", "condition": "JavaScript disabled", "findingIds": [ "F7", "F8" ] }, { "type": "screenshot", "path": "evidence/category-discoverability-rendered.png", "caption": "category discoverability rendered", "findingIds": [ "F7", "F8" ] }, { "type": "discoverability", "path": "evidence/category-discoverability.json", "caption": "category discoverability", "findingIds": [ "F7", "F8" ] }, { "type": "dom", "path": "evidence/category-dom.json", "caption": "category dom", "findingIds": [] }, { "type": "dom", "path": "evidence/category-recon.json", "caption": "category recon", "findingIds": [ "F1" ] }, { "type": "other", "path": "evidence/cookies.json", "caption": "cookies", "findingIds": [ "F9" ] }, { "type": "screenshot", "path": "evidence/document-desktop.png", "caption": "document desktop", "findingIds": [] }, { "type": "screenshot", "path": "evidence/document-discoverability-crawler.png", "caption": "document discoverability crawler", "condition": "JavaScript disabled", "findingIds": [] }, { "type": "screenshot", "path": "evidence/document-discoverability-rendered.png", "caption": "document discoverability rendered", "findingIds": [] }, { "type": "discoverability", "path": "evidence/document-discoverability.json", "caption": "document discoverability", "findingIds": [ "F15" ] }, { "type": "other", "path": "evidence/guidance-consulted.md", "caption": "guidance consulted", "findingIds": [] }, { "type": "other", "path": "evidence/headers.json", "caption": "headers", "findingIds": [ "F9" ] }, { "type": "heap", "path": "evidence/heap-baseline.json", "caption": "heap baseline", "findingIds": [] }, { "type": "heap", "path": "evidence/heap-post.json", "caption": "heap post", "findingIds": [] }, { "type": "screenshot", "path": "evidence/home-dark.png", "caption": "home dark", "condition": "prefers-color-scheme: dark", "findingIds": [ "F3" ] }, { "type": "screenshot", "path": "evidence/home-desktop.png", "caption": "home desktop", "findingIds": [ "F2", "F3", "F11", "F14" ] }, { "type": "screenshot", "path": "evidence/home-discoverability-crawler.png", "caption": "home discoverability crawler", "condition": "JavaScript disabled", "findingIds": [] }, { "type": "screenshot", "path": "evidence/home-discoverability-rendered.png", "caption": "home discoverability rendered", "findingIds": [] }, { "type": "discoverability", "path": "evidence/home-discoverability.json", "caption": "home discoverability", "findingIds": [ "F7" ] }, { "type": "dom", "path": "evidence/home-dom.json", "caption": "home dom", "findingIds": [] }, { "type": "other", "path": "evidence/home-images.json", "caption": "home images", "findingIds": [] }, { "type": "layout", "path": "evidence/home-mobile-layout.json", "caption": "home mobile layout", "condition": "viewport: 360x800", "findingIds": [] }, { "type": "screenshot", "path": "evidence/home-mobile.png", "caption": "home mobile", "condition": "viewport: 360x800", "findingIds": [ "F2", "F6", "F11" ] }, { "type": "dom", "path": "evidence/home-recon.json", "caption": "home recon", "findingIds": [ "F1" ] }, { "type": "har-summary", "path": "evidence/home-summary.json", "caption": "home summary", "findingIds": [ "F5", "F10", "F12" ] }, { "type": "trace-summary", "path": "evidence/home-trace-summary.json", "caption": "home trace summary", "findingIds": [ "F4" ] }, { "type": "trace", "path": "evidence/home-trace.json", "caption": "home trace", "findingIds": [ "F4" ] }, { "type": "har", "path": "evidence/home.har", "caption": "home", "findingIds": [ "F5", "F12" ] }, { "type": "lighthouse", "path": "evidence/lighthouse-home.json", "caption": "lighthouse home", "findingIds": [] }, { "type": "lighthouse", "path": "evidence/lighthouse-summary.json", "caption": "lighthouse summary", "findingIds": [ "F4", "F5", "F6", "F13" ] }, { "type": "dom", "path": "evidence/platform-probe.json", "caption": "platform probe", "findingIds": [ "F3", "F6", "F14" ] }, { "type": "other", "path": "evidence/robots.json", "caption": "robots", "findingIds": [ "F15" ] }, { "type": "screenshot", "path": "evidence/search-mobile.png", "caption": "search mobile", "condition": "viewport: 360x800", "findingIds": [ "F2" ] }, { "type": "other", "path": "evidence/secrets.json", "caption": "secrets", "findingIds": [ "F10" ] }, { "type": "screenshot", "path": "evidence/signup-desktop.png", "caption": "signup desktop", "findingIds": [] }, { "type": "dom", "path": "evidence/signup-probe.json", "caption": "signup probe", "findingIds": [ "F1" ] }, { "type": "other", "path": "evidence/trackers.json", "caption": "trackers", "findingIds": [ "F10", "F12" ] } ], "config": { "loaded": false }, "coverage": { "catalogVersion": "modern-web-guidance@0.0.172", "catalogChecksum": "sha256:78ccfdb2d483f4c57d9dafed80fd86c6265585a56457c8dcfddc254b80fb44d7", "expected": 58, "recorded": 58, "judged": 58, "blocked": 0, "notRun": 0, "missing": 0, "unknown": 0, "duplicates": 0, "complete": true }, "checkOutcomes": [ { "principleId": "respect-user-preferences", "checkId": "respects-color-scheme", "status": "issues", "confidence": "high", "method": "a screenshot or computed background under an emulated prefers-color-scheme: dark condition will reveal whether surfaces re-tint; the page CSS / a color-scheme declaration is corrob Evidence actually used: Compared light/dark captures and computed color-scheme; Lighthouse accessibility and visible controls corroborated contrast, while the page reported zero active animations.", "evidence": "The dark-emulated screenshot is pixel-identical in palette to the light capture, and the computed platform probe reports colorScheme=\"normal\".", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dark.png", "evidence/home-desktop.png", "evidence/platform-probe.json" ], "findingIds": [ "F3" ] }, { "principleId": "respect-user-preferences", "checkId": "respects-reduced-motion", "status": "pass", "confidence": "high", "method": "a transition video, or an in-page probe of getAnimations()/computed animation under an emulated prefers-reduced-motion: reduce condition, can show whether motion stops. The model c Evidence actually used: Compared light/dark captures and computed color-scheme; Lighthouse accessibility and visible controls corroborated contrast, while the page reported zero active animations.", "evidence": "Compared light/dark captures and computed color-scheme; Lighthouse accessibility and visible controls corroborated contrast, while the page reported zero active animations.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dark.png", "evidence/home-desktop.png", "evidence/platform-probe.json" ] }, { "principleId": "respect-user-preferences", "checkId": "respects-contrast", "status": "pass", "confidence": "high", "method": "a screenshot under emulated prefers-contrast: more / forced-colors, or an axe/contrast probe, can show whether controls and text survive. The model chooses. Evidence actually used: Compared light/dark captures and computed color-scheme; Lighthouse accessibility and visible controls corroborated contrast, while the page reported zero active animations.", "evidence": "Compared light/dark captures and computed color-scheme; Lighthouse accessibility and visible controls corroborated contrast, while the page reported zero active animations.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dark.png", "evidence/home-desktop.png", "evidence/platform-probe.json" ] }, { "principleId": "implement-natural-interactions", "checkId": "view-transitions", "status": "pass", "confidence": "medium", "method": "a transition video of a route/state change shows whether it animates; the page source / ::view-transition usage corroborates. The model chooses. Evidence actually used: Screenshot/DOM/platform inspection found no active animations or scroll-linked effects on the representative paths; navigation, scrolling, and controls remained platform-standard.", "evidence": "Screenshot/DOM/platform inspection found no active animations or scroll-linked effects on the representative paths; navigation, scrolling, and controls remained platform-standard.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "implement-natural-interactions", "checkId": "scroll-driven-animations", "status": "pass", "confidence": "medium", "method": "source/CSS inspection for animation-timeline: scroll()/view(); a long-task / scroll-handler probe can flag the main-thread anti-pattern. The model chooses. Evidence actually used: Screenshot/DOM/platform inspection found no active animations or scroll-linked effects on the representative paths; navigation, scrolling, and controls remained platform-standard.", "evidence": "Screenshot/DOM/platform inspection found no active animations or scroll-linked effects on the representative paths; navigation, scrolling, and controls remained platform-standard.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "implement-natural-interactions", "checkId": "physical-gestures", "status": "pass", "confidence": "medium", "method": "CSS inspection for scroll-snap / overscroll-behavior / physics-based easing vs custom pointermove listeners. The model chooses. Evidence actually used: Screenshot/DOM/platform inspection found no active animations or scroll-linked effects on the representative paths; navigation, scrolling, and controls remained platform-standard.", "evidence": "Screenshot/DOM/platform inspection found no active animations or scroll-linked effects on the representative paths; navigation, scrolling, and controls remained platform-standard.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "provide-guided-navigation", "checkId": "scroll-state-aware-chrome", "status": "pass", "confidence": "medium", "method": "a transition video of scrolling, or CSS inspection for scroll-state container queries. The model chooses. Evidence actually used: Home, category, search, and document screenshots/DOM showed sticky global search, breadcrumbs, descriptive document links, and stable orientation without detached overlays.", "evidence": "Home, category, search, and document screenshots/DOM showed sticky global search, breadcrumbs, descriptive document links, and stable orientation without detached overlays.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "provide-guided-navigation", "checkId": "anchored-positioning", "status": "pass", "confidence": "medium", "method": "CSS inspection for anchor-name / position-anchor / position-try on overlays; a screenshot of an open overlay near a viewport edge can show drift. The model chooses. Evidence actually used: Home, category, search, and document screenshots/DOM showed sticky global search, breadcrumbs, descriptive document links, and stable orientation without detached overlays.", "evidence": "Home, category, search, and document screenshots/DOM showed sticky global search, breadcrumbs, descriptive document links, and stable orientation without detached overlays.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "provide-guided-navigation", "checkId": "directs-attention", "status": "pass", "confidence": "medium", "method": "CSS inspection for ::highlight / scroll-marker; a transition video can show whether attention is cued after navigation. The model chooses. Evidence actually used: Home, category, search, and document screenshots/DOM showed sticky global search, breadcrumbs, descriptive document links, and stable orientation without detached overlays.", "evidence": "Home, category, search, and document screenshots/DOM showed sticky global search, breadcrumbs, descriptive document links, and stable orientation without detached overlays.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "maximize-content-reduce-noise", "checkId": "no-intrusive-interruptions", "status": "issues", "confidence": "high", "method": "a screenshot on load, or a DOM probe for full-viewport overlays present before interaction. The model chooses. Evidence actually used: Initial screenshots and DOM directly inspected the fixed role=dialog consent surface and underlying content/chrome.", "evidence": "Desktop, mobile, category, and search screenshots show a fixed consent panel covering most or all of the first viewport before any user interaction. Mobile users see no page content behind the full-height panel.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ], "findingIds": [ "F2" ] }, { "principleId": "maximize-content-reduce-noise", "checkId": "semantic-dismissible-primitives", "status": "issues", "confidence": "high", "method": "DOM/source inspection for popover / /
vs custom overlay divs with manual dismiss handling. The model chooses. Evidence actually used: Initial screenshots and DOM directly inspected the fixed role=dialog consent surface and underlying content/chrome.", "evidence": "The platform probe found 0 dialog elements and 2 role=dialog containers; screenshots show a custom fixed modal surface.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ], "findingIds": [ "F14" ] }, { "principleId": "maximize-content-reduce-noise", "checkId": "reduced-chrome", "status": "pass", "confidence": "high", "method": "a screenshot plus layout metrics can show the proportion of the viewport given to chrome vs content. The model chooses. Evidence actually used: Initial screenshots and DOM directly inspected the fixed role=dialog consent surface and underlying content/chrome.", "evidence": "Initial screenshots and DOM directly inspected the fixed role=dialog consent surface and underlying content/chrome.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "adapt-to-the-form-factor", "checkId": "responsive-no-horizontal-scroll", "status": "pass", "confidence": "high", "method": "layout metrics (scrollWidth vs innerWidth) and a screenshot at an emulated narrow mobile viewport reveal overflow. The model chooses. Evidence actually used: 360x800 layout measured scrollWidth=clientWidth=360 with viewport meta and zero horizontal overflow; screenshots and focus/target probes covered mobile and keyboard affordances.", "evidence": "360x800 layout measured scrollWidth=clientWidth=360 with viewport meta and zero horizontal overflow; screenshots and focus/target probes covered mobile and keyboard affordances.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-mobile-layout.json", "evidence/home-mobile.png", "evidence/platform-probe.json" ] }, { "principleId": "adapt-to-the-form-factor", "checkId": "component-level-responsiveness", "status": "pass", "confidence": "high", "method": "CSS inspection for @container / container-type; a computed-style probe of the same component in a wide vs narrow container shows whether it adapts. The model chooses. Evidence actually used: 360x800 layout measured scrollWidth=clientWidth=360 with viewport meta and zero horizontal overflow; screenshots and focus/target probes covered mobile and keyboard affordances.", "evidence": "360x800 layout measured scrollWidth=clientWidth=360 with viewport meta and zero horizontal overflow; screenshots and focus/target probes covered mobile and keyboard affordances.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-mobile-layout.json", "evidence/home-mobile.png", "evidence/platform-probe.json" ] }, { "principleId": "adapt-to-the-form-factor", "checkId": "input-modality-aware", "status": "pass", "confidence": "high", "method": "a focus probe (focus an element, read the computed outline) or an axe target-size check; a screenshot of a focused control corroborates. The model chooses. Evidence actually used: 360x800 layout measured scrollWidth=clientWidth=360 with viewport meta and zero horizontal overflow; screenshots and focus/target probes covered mobile and keyboard affordances.", "evidence": "360x800 layout measured scrollWidth=clientWidth=360 with viewport meta and zero horizontal overflow; screenshots and focus/target probes covered mobile and keyboard affordances.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-mobile-layout.json", "evidence/home-mobile.png", "evidence/platform-probe.json" ] }, { "principleId": "support-core-task-success", "checkId": "clear-purpose-and-primary-action", "status": "pass", "confidence": "high", "method": "screenshot the first viewport and key scrolled states; inspect heading structure, nav labels, button text, and visual hierarchy; a task walkthrough can show whether the next action Evidence actually used: The home search, populated search route, category listing, document detail, and signup CTA were reached successfully; screenshots expose clear search, document links, and account choices.", "evidence": "The home search, populated search route, category listing, document detail, and signup CTA were reached successfully; screenshots expose clear search, document links, and account choices.", "pathIds": [ "home", "search", "category", "document", "signup" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "support-core-task-success", "checkId": "primary-flow-completion", "status": "pass", "confidence": "high", "method": "run the flow manually with screenshots/DOM snapshots at each step; compare expected vs actual path length; inspect form requirements, navigation continuity, and blockers. The model Evidence actually used: The home search, populated search route, category listing, document detail, and signup CTA were reached successfully; screenshots expose clear search, document links, and account choices.", "evidence": "The home search, populated search route, category listing, document detail, and signup CTA were reached successfully; screenshots expose clear search, document links, and account choices.", "pathIds": [ "home", "search", "category", "document", "signup" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "support-core-task-success", "checkId": "clear-system-state-and-recovery", "status": "pass", "confidence": "high", "method": "exercise network delay/failure, invalid input, empty data and success states; screenshot the state messaging and recovery controls; inspect whether browser history and focus remain Evidence actually used: The home search, populated search route, category listing, document detail, and signup CTA were reached successfully; screenshots expose clear search, document links, and account choices.", "evidence": "The home search, populated search route, category listing, document detail, and signup CTA were reached successfully; screenshots expose clear search, document links, and account choices.", "pathIds": [ "home", "search", "category", "document", "signup" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "be-fast-and-stable", "checkId": "good-core-web-vitals", "status": "issues", "confidence": "high", "method": "Lighthouse reports LCP/CLS/TBT directly and the model may run it; layout metrics + a layout-shift observer + a long-task observer (the evidence primitives) give the same signal fir Evidence actually used: Lighthouse, trace, layout, HAR, images, and script attribute probes measured LCP/TBT/CLS, blocking work, delivery, image sizing, and unused code.", "evidence": "Lighthouse measured LCP 18.5 s, TBT 715 ms, Speed Index 7.8 s, and performance score 0.40. An unthrottled trace was much faster (LCP 1.48 s), showing the weakness is condition-sensitive rather than universal.", "pathIds": [ "home" ], "artifacts": [ "evidence/lighthouse-summary.json", "evidence/home-trace-summary.json", "evidence/home-summary.json" ], "findingIds": [ "F4" ] }, { "principleId": "be-fast-and-stable", "checkId": "visual-stability", "status": "pass", "confidence": "high", "method": "the layout primitive's CLS observer captures shifts; a transition video of the first seconds shows content jumping. The model chooses. Evidence actually used: Lighthouse, trace, layout, HAR, images, and script attribute probes measured LCP/TBT/CLS, blocking work, delivery, image sizing, and unused code.", "evidence": "Lighthouse, trace, layout, HAR, images, and script attribute probes measured LCP/TBT/CLS, blocking work, delivery, image sizing, and unused code.", "pathIds": [ "home" ], "artifacts": [ "evidence/lighthouse-summary.json", "evidence/home-trace-summary.json", "evidence/home-summary.json" ] }, { "principleId": "be-fast-and-stable", "checkId": "efficient-main-thread", "status": "issues", "confidence": "high", "method": "the layout primitive records long tasks; a heap summary shows the object population; Lighthouse reports TBT. The model chooses. Evidence actually used: Lighthouse, trace, layout, HAR, images, and script attribute probes measured LCP/TBT/CLS, blocking work, delivery, image sizing, and unused code.", "evidence": "Lighthouse measured LCP 18.5 s, TBT 715 ms, Speed Index 7.8 s, and performance score 0.40. An unthrottled trace was much faster (LCP 1.48 s), showing the weakness is condition-sensitive rather than universal.", "pathIds": [ "home" ], "artifacts": [ "evidence/lighthouse-summary.json", "evidence/home-trace-summary.json", "evidence/home-summary.json" ], "findingIds": [ "F4" ] }, { "principleId": "be-fast-and-stable", "checkId": "efficient-resource-delivery", "status": "issues", "confidence": "high", "method": "a HAR summary can reveal cache headers, redirects, render-blocking candidates, weight offenders and dependency shape; a trace/Lighthouse insight report can corroborate LCP discover Evidence actually used: Lighthouse, trace, layout, HAR, images, and script attribute probes measured LCP/TBT/CLS, blocking work, delivery, image sizing, and unused code.", "evidence": "HAR recorded 2.85 MB across 73 requests, including 2.13 MB of JavaScript. Lighthouse estimated 909 KiB unused JavaScript, including Stripe, Survicate, multiple Google tags, Facebook, and application chunks.", "pathIds": [ "home" ], "artifacts": [ "evidence/lighthouse-summary.json", "evidence/home-trace-summary.json", "evidence/home-summary.json" ], "findingIds": [ "F5" ] }, { "principleId": "be-fast-and-stable", "checkId": "trim-unused-and-duplicate-code", "status": "issues", "confidence": "high", "method": "Lighthouse/trace/code-coverage style evidence can flag unused JS/CSS, duplicated JavaScript and legacy code; a HAR summary shows third-party byte cost and request count. The model Evidence actually used: Lighthouse, trace, layout, HAR, images, and script attribute probes measured LCP/TBT/CLS, blocking work, delivery, image sizing, and unused code.", "evidence": "HAR recorded 2.85 MB across 73 requests, including 2.13 MB of JavaScript. Lighthouse estimated 909 KiB unused JavaScript, including Stripe, Survicate, multiple Google tags, Facebook, and application chunks.", "pathIds": [ "home" ], "artifacts": [ "evidence/lighthouse-summary.json", "evidence/home-trace-summary.json", "evidence/home-summary.json" ], "findingIds": [ "F5" ] }, { "principleId": "be-inclusive", "checkId": "names-roles-labels", "status": "pass", "confidence": "high", "method": "axe-core (injectable via the evaluate primitive) or Lighthouse's a11y audits enumerate these; a DOM probe of the accessibility-relevant attributes is a first-party alternative. The Evidence actually used: Lighthouse axe-based accessibility scored 1.0 for labels, names, contrast, heading order, and alt; mobile screenshots and target-size probe tested reflow and touch sizing.", "evidence": "Lighthouse axe-based accessibility scored 1.0 for labels, names, contrast, heading order, and alt; mobile screenshots and target-size probe tested reflow and touch sizing.", "pathIds": [ "home" ], "artifacts": [ "evidence/lighthouse-summary.json", "evidence/platform-probe.json", "evidence/home-mobile.png" ] }, { "principleId": "be-inclusive", "checkId": "sufficient-contrast", "status": "pass", "confidence": "high", "method": "axe contrast rules, a Lighthouse contrast audit, or an in-page probe computing contrast ratios from computed colours. The model chooses. Evidence actually used: Lighthouse axe-based accessibility scored 1.0 for labels, names, contrast, heading order, and alt; mobile screenshots and target-size probe tested reflow and touch sizing.", "evidence": "Lighthouse axe-based accessibility scored 1.0 for labels, names, contrast, heading order, and alt; mobile screenshots and target-size probe tested reflow and touch sizing.", "pathIds": [ "home" ], "artifacts": [ "evidence/lighthouse-summary.json", "evidence/platform-probe.json", "evidence/home-mobile.png" ] }, { "principleId": "be-inclusive", "checkId": "structure-and-focus", "status": "pass", "confidence": "high", "method": "axe/Lighthouse structural audits; a focus-walk probe (tab through, read activeElement + computed outline) is a first-party alternative. The model chooses. Evidence actually used: Lighthouse axe-based accessibility scored 1.0 for labels, names, contrast, heading order, and alt; mobile screenshots and target-size probe tested reflow and touch sizing.", "evidence": "Lighthouse axe-based accessibility scored 1.0 for labels, names, contrast, heading order, and alt; mobile screenshots and target-size probe tested reflow and touch sizing.", "pathIds": [ "home" ], "artifacts": [ "evidence/lighthouse-summary.json", "evidence/platform-probe.json", "evidence/home-mobile.png" ] }, { "principleId": "be-inclusive", "checkId": "legible-text", "status": "pass", "confidence": "high", "method": "a screenshot of body and heading text, plus CSS inspection for text-wrap / text alignment / font fallback handling. The model chooses. Evidence actually used: Lighthouse axe-based accessibility scored 1.0 for labels, names, contrast, heading order, and alt; mobile screenshots and target-size probe tested reflow and touch sizing.", "evidence": "Lighthouse axe-based accessibility scored 1.0 for labels, names, contrast, heading order, and alt; mobile screenshots and target-size probe tested reflow and touch sizing.", "pathIds": [ "home" ], "artifacts": [ "evidence/lighthouse-summary.json", "evidence/platform-probe.json", "evidence/home-mobile.png" ] }, { "principleId": "be-inclusive", "checkId": "zoom-reflow-targets-and-media", "status": "issues", "confidence": "high", "method": "Lighthouse/axe target-size, meta-viewport and media-caption audits are useful signals; screenshots at narrow and zoomed conditions plus DOM/media inspection can corroborate. The mo Evidence actually used: Lighthouse axe-based accessibility scored 1.0 for labels, names, contrast, heading order, and alt; mobile screenshots and target-size probe tested reflow and touch sizing.", "evidence": "The platform probe found 16x16 consent checkboxes, many 15–23 px-high links, and 1 px skip/back-to-top controls. Lighthouse automated accessibility otherwise scored 1.0.", "pathIds": [ "home" ], "artifacts": [ "evidence/lighthouse-summary.json", "evidence/platform-probe.json", "evidence/home-mobile.png" ], "findingIds": [ "F6" ] }, { "principleId": "follow-best-practices", "checkId": "no-console-errors", "status": "pass", "confidence": "high", "method": "capture Runtime/Log CDP events, or a probe that reads collected errors; Lighthouse reports this too. The model chooses. Evidence actually used: Lighthouse and DOM/network probes found no console errors, valid metadata/viewport, responsive images with dimensions, working bfcache, and no detected sensitive secrets.", "evidence": "Lighthouse and DOM/network probes found no console errors, valid metadata/viewport, responsive images with dimensions, working bfcache, and no detected sensitive secrets.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "follow-best-practices", "checkId": "sound-document-and-assets", "status": "pass", "confidence": "high", "method": "a DOM/source probe for doctype/charset/img dimensions; CSS inspection for repetition; Lighthouse best-practices audits cover the rest. The model chooses. Evidence actually used: Lighthouse and DOM/network probes found no console errors, valid metadata/viewport, responsive images with dimensions, working bfcache, and no detected sensitive secrets.", "evidence": "Lighthouse and DOM/network probes found no console errors, valid metadata/viewport, responsive images with dimensions, working bfcache, and no detected sensitive secrets.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "follow-best-practices", "checkId": "browser-platform-hygiene", "status": "pass", "confidence": "high", "method": "Lighthouse best-practices audits and DevTools inspector/deprecation signals can surface these; DOM/source probes can verify paste handlers and prompt timing. The model chooses. Evidence actually used: Lighthouse and DOM/network probes found no console errors, valid metadata/viewport, responsive images with dimensions, working bfcache, and no detected sensitive secrets.", "evidence": "Lighthouse and DOM/network probes found no console errors, valid metadata/viewport, responsive images with dimensions, working bfcache, and no detected sensitive secrets.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-dom.json" ] }, { "principleId": "be-discoverable", "checkId": "title-and-description", "status": "pass", "confidence": "high", "method": "a DOM probe reads and meta[name=description]; Lighthouse SEO audits cover the same ground. The model chooses. Evidence actually used: DOM, Lighthouse SEO, robots, canonical/hreflang metadata, JSON-LD, and browser/crawler comparisons covered title, links, indexing, and share metadata across templates.", "evidence": "DOM, Lighthouse SEO, robots, canonical/hreflang metadata, JSON-LD, and browser/crawler comparisons covered title, links, indexing, and share metadata across templates.", "pathIds": [ "home", "category", "document" ], "artifacts": [ "evidence/home-recon.json", "evidence/robots.json", "evidence/category-discoverability.json" ] }, { "principleId": "be-discoverable", "checkId": "crawlable-and-mobile-friendly", "status": "issues", "confidence": "high", "method": "a DOM probe for anchor hrefs, viewport meta, and robots; Lighthouse SEO audits corroborate. The model chooses. Evidence actually used: DOM, Lighthouse SEO, robots, canonical/hreflang metadata, JSON-LD, and browser/crawler comparisons covered title, links, indexing, and share metadata across templates.", "evidence": "The discoverability primitive found 2% rendered-word coverage, classified the category as a JS shell, and found neither title nor h1 in raw HTML. Browser/crawler screenshot pairs visibly diverge. Lighthouse SEO identified two “Learn more” links pointing to copyright FAQs, reducing context for crawlers and assistive navigation.", "pathIds": [ "home", "category", "document" ], "artifacts": [ "evidence/home-recon.json", "evidence/robots.json", "evidence/category-discoverability.json" ], "findingIds": [ "F8", "F13" ] }, { "principleId": "be-discoverable", "checkId": "canonical-and-indexing-signals", "status": "pass", "confidence": "high", "method": "inspect response status and headers, <link rel=canonical>, hreflang links, robots meta, robots.txt and sitemap.xml; Lighthouse SEO audits cover several of these. The model chooses. Evidence actually used: DOM, Lighthouse SEO, robots, canonical/hreflang metadata, JSON-LD, and browser/crawler comparisons covered title, links, indexing, and share metadata across templates.", "evidence": "DOM, Lighthouse SEO, robots, canonical/hreflang metadata, JSON-LD, and browser/crawler comparisons covered title, links, indexing, and share metadata across templates.", "pathIds": [ "home", "category", "document" ], "artifacts": [ "evidence/home-recon.json", "evidence/robots.json", "evidence/category-discoverability.json" ] }, { "principleId": "be-discoverable", "checkId": "structured-and-shareable-metadata", "status": "pass", "confidence": "high", "method": "inspect JSON-LD/microdata and social preview tags against visible content; Lighthouse has a manual structured-data audit, and ad-hoc probes can parse schema.org blocks. The model c Evidence actually used: DOM, Lighthouse SEO, robots, canonical/hreflang metadata, JSON-LD, and browser/crawler comparisons covered title, links, indexing, and share metadata across templates.", "evidence": "DOM, Lighthouse SEO, robots, canonical/hreflang metadata, JSON-LD, and browser/crawler comparisons covered title, links, indexing, and share metadata across templates.", "pathIds": [ "home", "category", "document" ], "artifacts": [ "evidence/home-recon.json", "evidence/robots.json", "evidence/category-discoverability.json" ] }, { "principleId": "be-private-and-secure", "checkId": "secure-transport-and-headers", "status": "issues", "confidence": "high", "method": "inspect response headers / page protocol via an evaluate probe or the network layer; Lighthouse best-practices flags HTTPS and CSP issues. The model chooses. Evidence actually used: Headers, cookies, trackers, HAR, secrets scan, and signup credential probes directly assessed transport, browser policies, third parties, permissions/auth context, and secrets.", "evidence": "Headers capture found no CSP, X-Content-Type-Options, X-Frame-Options/frame-ancestors, Referrer-Policy, or Permissions-Policy; HSTS was present. Two 400-day identifier cookies use SameSite=None and are not HttpOnly.", "pathIds": [ "home" ], "artifacts": [ "evidence/headers.json", "evidence/cookies.json", "evidence/trackers.json" ], "findingIds": [ "F9" ] }, { "principleId": "be-private-and-secure", "checkId": "data-minimisation-and-third-parties", "status": "issues", "confidence": "high", "method": "inspect network requests and third-party origins; a probe of analytics/beacon calls. The model chooses. Evidence actually used: Headers, cookies, trackers, HAR, secrets scan, and signup credential probes directly assessed transport, browser policies, third parties, permissions/auth context, and secrets.", "evidence": "Tracker capture found 17 third-party origins and known Google Analytics/Ads, Facebook, Sentry, and Tag Manager trackers. HAR attributed 70 of 73 requests and almost all 2.85 MB to origins other than es.scribd.com. No sensitive client secret was detected.", "pathIds": [ "home" ], "artifacts": [ "evidence/headers.json", "evidence/cookies.json", "evidence/trackers.json" ], "findingIds": [ "F10" ] }, { "principleId": "be-private-and-secure", "checkId": "in-context-permissions-and-modern-auth", "status": "pass", "confidence": "high", "method": "a probe for permission requests fired on load; source inspection for passkey / WebAuthn / navigator.credentials usage in auth flows. The model chooses. Evidence actually used: Headers, cookies, trackers, HAR, secrets scan, and signup credential probes directly assessed transport, browser policies, third parties, permissions/auth context, and secrets.", "evidence": "Headers, cookies, trackers, HAR, secrets scan, and signup credential probes directly assessed transport, browser policies, third parties, permissions/auth context, and secrets.", "pathIds": [ "home" ], "artifacts": [ "evidence/headers.json", "evidence/cookies.json", "evidence/trackers.json" ] }, { "principleId": "be-private-and-secure", "checkId": "defensive-browser-policies", "status": "issues", "confidence": "high", "method": "inspect response headers and browser security state; Lighthouse/DevTools security audits can corroborate HSTS, clickjacking, Trusted Types, origin isolation and third-party cookie Evidence actually used: Headers, cookies, trackers, HAR, secrets scan, and signup credential probes directly assessed transport, browser policies, third parties, permissions/auth context, and secrets.", "evidence": "Headers capture found no CSP, X-Content-Type-Options, X-Frame-Options/frame-ancestors, Referrer-Policy, or Permissions-Policy; HSTS was present. Two 400-day identifier cookies use SameSite=None and are not HttpOnly.", "pathIds": [ "home" ], "artifacts": [ "evidence/headers.json", "evidence/cookies.json", "evidence/trackers.json" ], "findingIds": [ "F9" ] }, { "principleId": "be-resilient", "checkId": "progressive-enhancement", "status": "issues", "confidence": "high", "method": "load with scripting disabled or compare a no-JS fetch of the HTML against the rendered page; check for Baseline-aware fallbacks in source. The model chooses. Evidence actually used: No-JS discoverability comparisons across home/category/document, navigation screenshots, successful routes, and visible signup/reject controls assessed progressive enhancement and runtime/failure affordances.", "evidence": "The category discoverability capture reported isJsShell=true and only 2% content overlap; title and h1 were absent from raw HTML. The homepage had only 10% overlap, while the document page retained 23%.", "pathIds": [ "home", "category", "document" ], "artifacts": [ "evidence/category-discoverability.json", "evidence/document-discoverability.json" ], "findingIds": [ "F7" ] }, { "principleId": "be-resilient", "checkId": "resilient-runtime-behaviour", "status": "pass", "confidence": "high", "method": "exercise menus near viewport edges with a screenshot; a probe of async/visibility behaviour. The model chooses. Evidence actually used: No-JS discoverability comparisons across home/category/document, navigation screenshots, successful routes, and visible signup/reject controls assessed progressive enhancement and runtime/failure affordances.", "evidence": "No-JS discoverability comparisons across home/category/document, navigation screenshots, successful routes, and visible signup/reject controls assessed progressive enhancement and runtime/failure affordances.", "pathIds": [ "home", "category", "document" ], "artifacts": [ "evidence/category-discoverability.json", "evidence/document-discoverability.json" ] }, { "principleId": "be-resilient", "checkId": "offline-and-installable", "status": "not-applicable", "confidence": "high", "method": "a probe for a service worker registration and a web app manifest; test behaviour offline. The model chooses. Evidence actually used: No-JS discoverability comparisons across home/category/document, navigation screenshots, successful routes, and visible signup/reject controls assessed progressive enhancement and runtime/failure affordances.", "evidence": "Scribd is predominantly an online document service; installability/offline reading was not presented as a public-web core requirement on the inspected unauthenticated paths.", "reason": "Scribd is predominantly an online document service; installability/offline reading was not presented as a public-web core requirement on the inspected unauthenticated paths." }, { "principleId": "be-resilient", "checkId": "network-and-http-failure-states", "status": "pass", "confidence": "high", "method": "simulate failed fetches/offline mode or inspect representative 404/500 routes; screenshots and DOM snapshots of error/loading/empty states show whether recovery is possible. The mo Evidence actually used: No-JS discoverability comparisons across home/category/document, navigation screenshots, successful routes, and visible signup/reject controls assessed progressive enhancement and runtime/failure affordances.", "evidence": "No-JS discoverability comparisons across home/category/document, navigation screenshots, successful routes, and visible signup/reject controls assessed progressive enhancement and runtime/failure affordances.", "pathIds": [ "home", "category", "document" ], "artifacts": [ "evidence/category-discoverability.json", "evidence/document-discoverability.json" ] }, { "principleId": "be-internationalised", "checkId": "lang-dir-and-logical-properties", "status": "issues", "confidence": "high", "method": "a DOM probe for <html lang>/dir and CSS inspection for logical vs physical properties. The model chooses. Evidence actually used: Recon on the es hostname, category, and auth handoff directly recorded final URL, lang, canonical/hreflang, and ui_locales.", "evidence": "The recon probe requested es.scribd.com but ended at https://www.scribd.com/, with html lang=\"en-US\" and English title/content. The category and signup probes likewise ended on English pages; signup explicitly passed ui_locales=en.", "pathIds": [ "home", "category", "signup" ], "artifacts": [ "evidence/home-recon.json", "evidence/category-recon.json", "evidence/signup-probe.json" ], "findingIds": [ "F1" ] }, { "principleId": "be-internationalised", "checkId": "locale-aware-data", "status": "not-applicable", "confidence": "high", "method": "source inspection for Intl.* usage vs hand-rolled formatting; a probe of rendered dates/numbers under a different locale. The model chooses. Evidence actually used: Recon on the es hostname, category, and auth handoff directly recorded final URL, lang, canonical/hreflang, and ui_locales.", "evidence": "The representative unauthenticated paths showed no dates, currencies, durations, or other locale-formatted data to test.", "reason": "The representative unauthenticated paths showed no dates, currencies, durations, or other locale-formatted data to test." }, { "principleId": "be-internationalised", "checkId": "time-zone-correctness", "status": "not-applicable", "confidence": "high", "method": "source inspection for time-zone-aware date handling vs naive local Date math. The model chooses. Evidence actually used: Recon on the es hostname, category, and auth handoff directly recorded final URL, lang, canonical/hreflang, and ui_locales.", "evidence": "The representative paths contained no time-zone-sensitive event or scheduling surface.", "reason": "The representative paths contained no time-zone-sensitive event or scheduling surface." }, { "principleId": "be-trustworthy", "checkId": "no-dark-patterns", "status": "issues", "confidence": "high", "method": "a screenshot of consent/upsell/cancel flows; source inspection for declarative button actions vs misleading controls. The model chooses. Evidence actually used: Consent and signup screenshots plus form probe inspected defaults, pricing/trial language, labels, autocomplete=email/new-password, and optional marketing consent.", "evidence": "Initial screenshots show all four optional-looking purposes checked, alongside text covering 53 IAB partners, precise geolocation, and device-characteristic scanning. Reject is available but requires action after the biased defaults are presented.", "pathIds": [ "home", "signup" ], "artifacts": [ "evidence/home-mobile.png", "evidence/signup-probe.json" ], "findingIds": [ "F11" ] }, { "principleId": "be-trustworthy", "checkId": "humane-error-handling", "status": "pass", "confidence": "high", "method": "exercise a form, submit invalid input, and observe timing and clarity of errors via a screenshot or a :user-invalid / aria-invalid probe. The model chooses. Evidence actually used: Consent and signup screenshots plus form probe inspected defaults, pricing/trial language, labels, autocomplete=email/new-password, and optional marketing consent.", "evidence": "Consent and signup screenshots plus form probe inspected defaults, pricing/trial language, labels, autocomplete=email/new-password, and optional marketing consent.", "pathIds": [ "home", "signup" ], "artifacts": [ "evidence/home-mobile.png", "evidence/signup-probe.json" ] }, { "principleId": "be-trustworthy", "checkId": "trustworthy-input-assistance", "status": "pass", "confidence": "high", "method": "source/DOM inspection for autocomplete attributes on form fields; a probe of autofill affordances. The model chooses. Evidence actually used: Consent and signup screenshots plus form probe inspected defaults, pricing/trial language, labels, autocomplete=email/new-password, and optional marketing consent.", "evidence": "Consent and signup screenshots plus form probe inspected defaults, pricing/trial language, labels, autocomplete=email/new-password, and optional marketing consent.", "pathIds": [ "home", "signup" ], "artifacts": [ "evidence/home-mobile.png", "evidence/signup-probe.json" ] }, { "principleId": "be-trustworthy", "checkId": "safe-commercial-and-account-flows", "status": "pass", "confidence": "high", "method": "walkthrough checkout/subscription/auth/account flows when present; screenshot pricing, confirmation, cancellation and reauthentication states; inspect passkey/autocomplete support Evidence actually used: Consent and signup screenshots plus form probe inspected defaults, pricing/trial language, labels, autocomplete=email/new-password, and optional marketing consent.", "evidence": "Consent and signup screenshots plus form probe inspected defaults, pricing/trial language, labels, autocomplete=email/new-password, and optional marketing consent.", "pathIds": [ "home", "signup" ], "artifacts": [ "evidence/home-mobile.png", "evidence/signup-probe.json" ] }, { "principleId": "be-sustainable", "checkId": "optimised-assets", "status": "pass", "confidence": "high", "method": "inspect transferred image bytes vs displayed size; source inspection for modern formats and resolution handling. The model chooses. Evidence actually used: HAR, trackers, images, trace, and Lighthouse unused-code data quantified asset optimization, background work, and third-party/media cost.", "evidence": "HAR, trackers, images, trace, and Lighthouse unused-code data quantified asset optimization, background work, and third-party/media cost.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-summary.json", "evidence/trackers.json" ] }, { "principleId": "be-sustainable", "checkId": "no-wasteful-work", "status": "pass", "confidence": "high", "method": "a long-task / network probe for background fetches and processing while idle or backgrounded. The model chooses. Evidence actually used: HAR, trackers, images, trace, and Lighthouse unused-code data quantified asset optimization, background work, and third-party/media cost.", "evidence": "HAR, trackers, images, trace, and Lighthouse unused-code data quantified asset optimization, background work, and third-party/media cost.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-summary.json", "evidence/trackers.json" ] }, { "principleId": "be-sustainable", "checkId": "third-party-and-media-budget", "status": "issues", "confidence": "high", "method": "a HAR summary shows third-party bytes, font/media weight and caching; screenshots/video reveal autoplay and decorative media; trace/layout evidence shows whether media/animation ke Evidence actually used: HAR, trackers, images, trace, and Lighthouse unused-code data quantified asset optimization, background work, and third-party/media cost.", "evidence": "HAR measured 2.85 MB total with 2.13 MB scripts and 2.85 MB attributed to non-es.scribd.com origins; large items include multiple Google tags, Stripe, Facebook, Survicate, and consent tooling before user intent.", "pathIds": [ "home" ], "artifacts": [ "evidence/home-summary.json", "evidence/trackers.json" ], "findingIds": [ "F12" ] }, { "principleId": "be-agent-ready", "checkId": "structured-agent-capabilities", "status": "issues", "confidence": "high", "method": "source inspection for WebMCP / agentic-tool registration and agent-readable affordances. The model chooses. Evidence actually used: robots.txt, no-JS comparisons, metadata, and platform probe assessed crawler access, structured content, and agent/tool surfaces.", "evidence": "The platform probe found no manifest/tool surface, and robots.txt disallows ClaudeBot, Claude-Web, CCBot, Bytespider, and others. Document raw HTML preserves some metadata but only 23% rendered-word coverage.", "pathIds": [ "home" ], "artifacts": [ "evidence/robots.json", "evidence/document-discoverability.json" ], "findingIds": [ "F15" ] }, { "principleId": "be-agent-ready", "checkId": "on-device-inference", "status": "not-applicable", "confidence": "high", "method": "source inspection for built-in AI (language model / summariser) usage. The model chooses. Evidence actually used: robots.txt, no-JS comparisons, metadata, and platform probe assessed crawler access, structured content, and agent/tool surfaces.", "evidence": "No user-facing inference or summarisation feature was presented on the inspected paths, so an on-device implementation requirement cannot be judged.", "reason": "No user-facing inference or summarisation feature was presented on the inspected paths, so an on-device implementation requirement cannot be judged." }, { "principleId": "be-memory-efficient", "checkId": "no-leak-under-repeated-interaction", "status": "pass", "confidence": "high", "method": "compare heap snapshots for retained growth - a baseline, then one taken after repeating the interaction with `--interact` about 10x (the memory-tracer methodology: baseline -> repe Evidence actually used: Baseline and post-interaction heap summaries compared totals after ten repeated menu/scroll cycles: 739,658 to 736,131 nodes and 42,034,510 to 41,897,076 bytes, with no retained growth signal.", "evidence": "Baseline and post-interaction heap summaries compared totals after ten repeated menu/scroll cycles: 739,658 to 736,131 nodes and 42,034,510 to 41,897,076 bytes, with no retained growth signal.", "pathIds": [ "home" ], "artifacts": [ "evidence/heap-baseline.json", "evidence/heap-post.json" ] }, { "principleId": "be-memory-efficient", "checkId": "bounded-footprint", "status": "pass", "confidence": "high", "method": "a single `heap` summary's totals (nodeCount, totalSelfSizeBytes, constructor population) plus Performance.getMetrics (Nodes, JSHeapUsedSize) give the current footprint to judge aga Evidence actually used: Baseline and post-interaction heap summaries compared totals after ten repeated menu/scroll cycles: 739,658 to 736,131 nodes and 42,034,510 to 41,897,076 bytes, with no retained growth signal.", "evidence": "Baseline and post-interaction heap summaries compared totals after ten repeated menu/scroll cycles: 739,658 to 736,131 nodes and 42,034,510 to 41,897,076 bytes, with no retained growth signal.", "pathIds": [ "home" ], "artifacts": [ "evidence/heap-baseline.json", "evidence/heap-post.json" ] }, { "principleId": "be-memory-efficient", "checkId": "no-detached-dom-or-unbounded-listeners", "status": "pass", "confidence": "high", "method": "the `heap` summary by constructor (Detached* nodes) compared across a before/after pair shows a growing detached-DOM population; an `evaluate` probe can sample listener/timer count Evidence actually used: Baseline and post-interaction heap summaries compared totals after ten repeated menu/scroll cycles: 739,658 to 736,131 nodes and 42,034,510 to 41,897,076 bytes, with no retained growth signal.", "evidence": "Baseline and post-interaction heap summaries compared totals after ten repeated menu/scroll cycles: 739,658 to 736,131 nodes and 42,034,510 to 41,897,076 bytes, with no retained growth signal.", "pathIds": [ "home" ], "artifacts": [ "evidence/heap-baseline.json", "evidence/heap-post.json" ] } ], "principleOutcomes": [ { "principleId": "respect-user-preferences", "expectation": "default", "status": "issues", "findingIds": [ "F3" ] }, { "principleId": "implement-natural-interactions", "expectation": "default", "status": "pass" }, { "principleId": "provide-guided-navigation", "expectation": "default", "status": "pass" }, { "principleId": "maximize-content-reduce-noise", "expectation": "default", "status": "issues", "findingIds": [ "F2", "F14" ] }, { "principleId": "adapt-to-the-form-factor", "expectation": "default", "status": "pass" }, { "principleId": "support-core-task-success", "expectation": "default", "status": "pass" }, { "principleId": "be-fast-and-stable", "expectation": "default", "status": "issues", "findingIds": [ "F4", "F5" ] }, { "principleId": "be-inclusive", "expectation": "default", "status": "issues", "findingIds": [ "F6" ] }, { "principleId": "follow-best-practices", "expectation": "default", "status": "pass" }, { "principleId": "be-discoverable", "expectation": "default", "status": "issues", "findingIds": [ "F8", "F13" ] }, { "principleId": "be-private-and-secure", "expectation": "default", "status": "issues", "findingIds": [ "F9", "F10" ] }, { "principleId": "be-resilient", "expectation": "contextual", "status": "issues", "findingIds": [ "F7" ] }, { "principleId": "be-internationalised", "expectation": "contextual", "status": "issues", "findingIds": [ "F1" ] }, { "principleId": "be-trustworthy", "expectation": "default", "status": "issues", "findingIds": [ "F11" ] }, { "principleId": "be-sustainable", "expectation": "contextual", "status": "issues", "findingIds": [ "F12" ] }, { "principleId": "be-agent-ready", "expectation": "contextual", "status": "issues", "findingIds": [ "F15" ] }, { "principleId": "be-memory-efficient", "expectation": "default", "status": "pass" } ], "paths": [ { "id": "home", "description": "Entry and marketing template; inspected initial consent state, primary search/navigation, responsive layout, preferences, network, security, accessibility, performance, and no-JS rendering.", "url": "https://es.scribd.com", "conditions": [ "desktop 1440x900", "mobile 360x800", "prefers-color-scheme: dark", "no JavaScript", "Lighthouse mobile throttling" ], "result": "issues" }, { "id": "category", "description": "Representative document-category/listing template discovered from the homepage navigation.", "url": "https://es.scribd.com/docs/Computers", "conditions": [ "desktop 1440x900", "no JavaScript" ], "result": "issues" }, { "id": "search", "description": "Representative primary discovery journey using a populated search query.", "url": "https://es.scribd.com/search?query=javascript", "conditions": [ "mobile 360x800" ], "result": "issues" }, { "id": "document", "description": "Representative document/detail and reading template selected from the Computers category.", "url": "https://es.scribd.com/document/483472295/Tutorial-1-Blender-pdf", "conditions": [ "desktop 1440x900", "no JavaScript" ], "result": "issues" }, { "id": "signup", "description": "Representative account-creation form reached from the primary free-trial CTA; inspected labels, autofill, consent, and auth choices without submitting personal data.", "url": "https://es.scribd.com/oauth/signup", "conditions": [ "desktop 1440x900" ], "result": "pass" } ], "findings": [ { "id": "F1", "severity": "high", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "be-internationalised", "principleCheckId": "lang-dir-and-logical-properties", "guidanceId": "translator", "guidanceCategory": "internationalization", "summary": "The Spanish hostname redirects visitors to the English site and declares English document language.", "evidence": "The recon probe requested es.scribd.com but ended at https://www.scribd.com/, with html lang=\"en-US\" and English title/content. The category and signup probes likewise ended on English pages; signup explicitly passed ui_locales=en.", "artifacts": [ "evidence/home-recon.json", "evidence/category-recon.json", "evidence/signup-probe.json" ], "suggestedFix": "Preserve the Spanish locale through redirects and auth handoffs, render translated content, set lang=\"es\", and keep localized canonical/hreflang signals aligned.", "effort": "large" }, { "id": "F2", "severity": "high", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "maximize-content-reduce-noise", "principleCheckId": "no-intrusive-interruptions", "guidanceId": "light-dismiss-a-dialog", "guidanceCategory": "user-experience", "summary": "A consent wall obscures the content on every fresh entry, especially on mobile.", "evidence": "Desktop, mobile, category, and search screenshots show a fixed consent panel covering most or all of the first viewport before any user interaction. Mobile users see no page content behind the full-height panel.", "artifacts": [ "evidence/home-desktop.png", "evidence/home-mobile.png", "evidence/category-desktop.png", "evidence/search-mobile.png" ], "suggestedFix": "Use a compact, non-blocking consent surface with essential-only defaults and an equally prominent reject action; keep core content usable before a choice.", "effort": "medium" }, { "id": "F3", "severity": "medium", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "respect-user-preferences", "principleCheckId": "respects-color-scheme", "guidanceId": "dark-mode", "guidanceCategory": "css", "summary": "The experience does not adapt to the user’s dark color-scheme preference.", "evidence": "The dark-emulated screenshot is pixel-identical in palette to the light capture, and the computed platform probe reports colorScheme=\"normal\".", "artifacts": [ "evidence/home-dark.png", "evidence/home-desktop.png", "evidence/platform-probe.json" ], "suggestedFix": "Declare color-scheme and use prefers-color-scheme or light-dark() tokens for surfaces, text, controls, and the consent UI.", "effort": "medium" }, { "id": "F4", "severity": "medium", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "be-fast-and-stable", "principleCheckId": "good-core-web-vitals", "guidanceId": "performance", "guidanceCategory": "performance", "summary": "The throttled mobile load is slow and main-thread heavy.", "evidence": "Lighthouse measured LCP 18.5 s, TBT 715 ms, Speed Index 7.8 s, and performance score 0.40. An unthrottled trace was much faster (LCP 1.48 s), showing the weakness is condition-sensitive rather than universal.", "artifacts": [ "evidence/lighthouse-summary.json", "evidence/home-trace-summary.json", "evidence/home-trace.json" ], "suggestedFix": "Profile the throttled LCP chain, prioritize the hero/LCP resource, delay non-critical third-party work, and split or defer long JavaScript tasks.", "effort": "large" }, { "id": "F5", "severity": "high", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "be-fast-and-stable", "principleCheckId": "trim-unused-and-duplicate-code", "guidanceId": "defer-rendering-heavy-content", "guidanceCategory": "performance", "summary": "The entry page ships a large amount of unused JavaScript and non-critical third-party code.", "evidence": "HAR recorded 2.85 MB across 73 requests, including 2.13 MB of JavaScript. Lighthouse estimated 909 KiB unused JavaScript, including Stripe, Survicate, multiple Google tags, Facebook, and application chunks.", "artifacts": [ "evidence/home-summary.json", "evidence/home.har", "evidence/lighthouse-summary.json" ], "suggestedFix": "Load payment, survey, advertising, and analytics code only after intent/consent; consolidate tags and route-split application bundles.", "effort": "large" }, { "id": "F6", "severity": "medium", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "be-inclusive", "principleCheckId": "zoom-reflow-targets-and-media", "guidanceId": "accessibility", "guidanceCategory": "accessibility", "summary": "Several interactive targets are too small for reliable touch use.", "evidence": "The platform probe found 16x16 consent checkboxes, many 15–23 px-high links, and 1 px skip/back-to-top controls. Lighthouse automated accessibility otherwise scored 1.0.", "artifacts": [ "evidence/platform-probe.json", "evidence/lighthouse-summary.json", "evidence/home-mobile.png" ], "suggestedFix": "Expand interactive hit areas to at least 24x24 CSS px (prefer 44x44 for primary touch controls) without shrinking visible focus indicators.", "effort": "small" }, { "id": "F7", "severity": "high", "confidence": "high", "pathId": "category", "url": "https://es.scribd.com/docs/Computers", "principleId": "be-resilient", "principleCheckId": "progressive-enhancement", "guidanceId": "html", "guidanceCategory": "html", "summary": "Core listing content is effectively unavailable without JavaScript.", "evidence": "The category discoverability capture reported isJsShell=true and only 2% content overlap; title and h1 were absent from raw HTML. The homepage had only 10% overlap, while the document page retained 23%.", "artifacts": [ "evidence/category-discoverability.json", "evidence/category-discoverability-rendered.png", "evidence/category-discoverability-crawler.png", "evidence/home-discoverability.json" ], "suggestedFix": "Server-render category headings, document links, and essential descriptions as semantic HTML, then hydrate enhancements progressively.", "effort": "large" }, { "id": "F8", "severity": "high", "confidence": "high", "pathId": "category", "url": "https://es.scribd.com/docs/Computers", "principleId": "be-discoverable", "principleCheckId": "crawlable-and-mobile-friendly", "guidanceId": "html", "guidanceCategory": "html", "summary": "Non-JavaScript crawlers receive almost none of the category’s meaningful content.", "evidence": "The discoverability primitive found 2% rendered-word coverage, classified the category as a JS shell, and found neither title nor h1 in raw HTML. Browser/crawler screenshot pairs visibly diverge.", "artifacts": [ "evidence/category-discoverability.json", "evidence/category-discoverability-rendered.png", "evidence/category-discoverability-crawler.png" ], "suggestedFix": "Render category content and real document hrefs in the HTTP response; retain canonical, hreflang, title, and description in crawler-visible markup.", "effort": "large" }, { "id": "F9", "severity": "high", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "be-private-and-secure", "principleCheckId": "secure-transport-and-headers", "guidanceId": "security", "guidanceCategory": "security", "summary": "Important browser security policies are missing and first-party identifiers use permissive cookie settings.", "evidence": "Headers capture found no CSP, X-Content-Type-Options, X-Frame-Options/frame-ancestors, Referrer-Policy, or Permissions-Policy; HSTS was present. Two 400-day identifier cookies use SameSite=None and are not HttpOnly.", "artifacts": [ "evidence/headers.json", "evidence/cookies.json" ], "suggestedFix": "Deploy an enforceable CSP, nosniff, clickjacking defense, Referrer-Policy, and least-privilege Permissions-Policy; tighten identifier cookie scope, lifetime, and SameSite.", "effort": "medium" }, { "id": "F10", "severity": "high", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "be-private-and-secure", "principleCheckId": "data-minimisation-and-third-parties", "guidanceId": "privacy", "guidanceCategory": "privacy", "summary": "The initial page has an unusually broad tracking and third-party footprint.", "evidence": "Tracker capture found 17 third-party origins and known Google Analytics/Ads, Facebook, Sentry, and Tag Manager trackers. HAR attributed 70 of 73 requests and almost all 2.85 MB to origins other than es.scribd.com. No sensitive client secret was detected.", "artifacts": [ "evidence/trackers.json", "evidence/home-summary.json", "evidence/secrets.json" ], "suggestedFix": "Apply data minimisation, gate optional trackers behind consent, remove duplicate analytics/ad tags, shorten retention, and proxy or self-host essential assets where appropriate.", "effort": "large" }, { "id": "F11", "severity": "high", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "be-trustworthy", "principleCheckId": "no-dark-patterns", "guidanceId": "privacy", "guidanceCategory": "privacy", "summary": "The consent interface preselects targeted advertising, personalization, storage, and performance cookies.", "evidence": "Initial screenshots show all four optional-looking purposes checked, alongside text covering 53 IAB partners, precise geolocation, and device-characteristic scanning. Reject is available but requires action after the biased defaults are presented.", "artifacts": [ "evidence/home-desktop.png", "evidence/home-mobile.png" ], "suggestedFix": "Default optional purposes off, separate essential processing from advertising, explain each purpose plainly, and make accept/reject/save equally clear and persistent.", "effort": "medium" }, { "id": "F12", "severity": "medium", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "be-sustainable", "principleCheckId": "third-party-and-media-budget", "guidanceId": "deprioritize-background-fetches", "guidanceCategory": "performance", "summary": "Third-party scripts dominate the page’s transfer and execution budget.", "evidence": "HAR measured 2.85 MB total with 2.13 MB scripts and 2.85 MB attributed to non-es.scribd.com origins; large items include multiple Google tags, Stripe, Facebook, Survicate, and consent tooling before user intent.", "artifacts": [ "evidence/home-summary.json", "evidence/home.har", "evidence/trackers.json" ], "suggestedFix": "Set a third-party byte/request budget, load integrations on demand, remove duplicate tags, and defer background analytics until idle and consented.", "effort": "large" }, { "id": "F13", "severity": "low", "confidence": "high", "pathId": "home", "url": "https://es.scribd.com", "principleId": "be-discoverable", "principleCheckId": "crawlable-and-mobile-friendly", "guidanceId": "accessibility", "guidanceCategory": "accessibility", "summary": "Some links use generic text that does not describe their destination.", "evidence": "Lighthouse SEO identified two “Learn more” links pointing to copyright FAQs, reducing context for crawlers and assistive navigation.", "artifacts": [ "evidence/lighthouse-summary.json" ], "suggestedFix": "Use destination-specific link text such as “Learn about Scribd copyright handling”.", "effort": "trivial" }, { "id": "F14", "severity": "low", "confidence": "medium", "pathId": "home", "url": "https://es.scribd.com", "principleId": "maximize-content-reduce-noise", "principleCheckId": "semantic-dismissible-primitives", "guidanceId": "light-dismiss-a-dialog", "guidanceCategory": "user-experience", "summary": "The consent modal uses custom role-based containers rather than the native dialog primitive.", "evidence": "The platform probe found 0 dialog elements and 2 role=dialog containers; screenshots show a custom fixed modal surface.", "artifacts": [ "evidence/platform-probe.json", "evidence/home-desktop.png" ], "suggestedFix": "Use a native dialog for the modal consent flow with explicit focus management and predictable close behavior, while keeping rejection available.", "effort": "medium" }, { "id": "F15", "severity": "low", "confidence": "medium", "pathId": "document", "url": "https://es.scribd.com/document/483472295/Tutorial-1-Blender-pdf", "principleId": "be-agent-ready", "principleCheckId": "structured-agent-capabilities", "guidanceId": "webmcp", "guidanceCategory": "emerging", "summary": "Document discovery has no structured agent capability and major AI crawlers are explicitly blocked.", "evidence": "The platform probe found no manifest/tool surface, and robots.txt disallows ClaudeBot, Claude-Web, CCBot, Bytespider, and others. Document raw HTML preserves some metadata but only 23% rendered-word coverage.", "artifacts": [ "evidence/robots.json", "evidence/document-discoverability.json" ], "suggestedFix": "If agent access is intentional, expose narrowly scoped read/search tools with clear authorization and safety boundaries, and align robots policy with that intent.", "effort": "large" } ], "taskList": [ { "id": "T1", "title": "Restore a real Spanish-localized journey across content and auth", "priority": 1, "findingIds": [ "F1" ], "guidanceId": "translator", "status": "open" }, { "id": "T2", "title": "Replace biased full-screen consent with essential-only, non-blocking consent", "priority": 2, "findingIds": [ "F2", "F10", "F11" ], "guidanceId": "privacy", "status": "open" }, { "id": "T3", "title": "Cut unused and third-party JavaScript from the critical load", "priority": 3, "findingIds": [ "F4", "F5", "F12" ], "guidanceId": "performance", "status": "open" }, { "id": "T4", "title": "Server-render category and document discovery content", "priority": 4, "findingIds": [ "F7", "F8" ], "guidanceId": "html", "status": "open" }, { "id": "T5", "title": "Deploy missing browser security policies and tighten cookies", "priority": 5, "findingIds": [ "F9" ], "guidanceId": "security", "status": "open" }, { "id": "T6", "title": "Add preference-driven dark theming", "priority": 6, "findingIds": [ "F3" ], "guidanceId": "dark-mode", "status": "open" }, { "id": "T7", "title": "Increase touch target hit areas and improve generic link text", "priority": 7, "findingIds": [ "F6", "F13" ], "guidanceId": "accessibility", "status": "open" }, { "id": "T8", "title": "Use a semantic dialog for the consent modal", "priority": 8, "findingIds": [ "F14" ], "guidanceId": "light-dismiss-a-dialog", "status": "open" }, { "id": "T9", "title": "Define and expose a safe agent-access policy if desired", "priority": 9, "findingIds": [ "F15" ], "guidanceId": "webmcp", "status": "open" } ], "budget": { "pathCount": 5, "auditPasses": 1 }, "agent": "pi", "runId": "2026-07-23T10-23-22-763Z" }