What’s new in the open beta

Added

• 🚀 Demo mode – run a timed Analyzer pack without creating an account and see example Overestimation/TLX tiles.
• 🔗 Intent-preserving auth – protected CTAs now append ?next= so you land on the exact analyzer view post-login.
• 🧭 Robots/AI guidelines – authored /robots.txt and /ai.txt so AI crawlers respect our usage policy.

Changed

• Navbar CTAs now adapt based on auth state and mention the demo when you are logged out.
• Login + signup screens detect ext and redirect accordingly.

Fixed

• Blocked access events now emit consistently, giving Ops a clean view of where guests get stuck.

Highlights

• 📊 Launched the literacy assessment + dashboard pair for early cohorts.
• ⚙️ Enabled analyzer task packs with manual vs. AI timing plus workload capture.
• 📚 Published the first wave of localized resources (EN + DE) with intent tagging.

Infrastructure

• Added Jest + Playwright quality gates for every PR.
• Provisioned nightly cron for psychometrics runs (now locked behind CRON_SECRET).

Added

• 🌐 i18n strings (EN + DE) – Three new translation keys added to lib/i18n/translations.ts:
  • analyzer.summaryTooltip – Explains composite score components
  • analyzer.judgmentPackIntro – Prediction-then-evaluation guidance
  • arena.optIn – Public leaderboard privacy notice
• ♿ Accessibility tests enabled – Re-enabled axe tests in e2e/accessibility.spec.ts covering Home, Login, Resources, Literacy Hub, Analyzer, and Arena pages
• ⌨️ Keyboard navigation test – Arena leaderboard filter buttons verified as keyboard accessible
• 🧪 EPIC F test coverage – New test file __tests__/epic-f-qa-gates.test.ts with 22 tests protecting:
  • i18n string presence and content (EN + DE)
  • Benchmark n<30 gate enforcement
  • Feed data structure validation
• 📡 Feed validation tests – New test file __tests__/feeds.test.ts with 13 tests verifying RSS feed data sources

Verified

• ✅ Randomization audit – Existing χ² tests in __tests__/quiz-randomization.test.ts pass (p > 0.05)
• ✅ Benchmark gate – lib/benchmarks.ts enforces MINIMUM_SAMPLE_SIZE = 30 before displaying percentile tiles
• ✅ SSR fallbacks – /analyzer redirects to /analyzer/noscript and /literacy-hub shows noscript notice when JS disabled
• ✅ RSS feeds – Both /feed.xml and /changelog.xml return 200 with valid RSS 2.0 structure

Test Results

• ESLint: 0 errors (3 warnings)
• TypeScript: Compiles successfully
• Unit tests: 254 passed (including 35 new tests)
• E2E accessibility: 10/12 tests pass (2 minor violations tracked)

Added

• 🧭 How it works anchors – The homepage now includes a diagram-driven “How it works” section plus shortcut link so visitors understand the loop in under a minute.
• 📊 My Results history – /my-results tracks assessment attempts, topic breakdowns, and filters so teams can revisit prior scores without digging through emails.
• 🧪 Randomization verification – Dedicated chi-squared unit tests (__tests__/quiz-randomization.test.ts) and helper scripts in /scripts ensure question/answer shuffle stays fair.
• 📰 Session documentation – Added docs/IMPLEMENTATION_SUMMARY.md + docs/SESSION_SUMMARY_NOV18.md to capture the sprint artifacts.

Changed

• 🏠 Homepage hero – New problem → solution → proof story with a single “Run the 3-minute demo” CTA, evidence highlights, and the newsletter moved below the fold.
• 🧾 Navigation – Primary nav now surfaces “Evidence” and a “How it works” jump link so the trust story is always one click away.
• 🧠 Analyzer SSR fallback – /analyzer/noscript keeps reporting functional for zero-JS environments.

Fixed

• ♿ Heading order – Adjusted the social-proof subheading to <h2> (app/page.tsx) so axe’s heading-order rule passes again during accessibility tests.
• 🔐 Protected flow clarity – The navbar strapline and CTA pairings now respect the updated hero narrative, reducing auth wall confusion.

Added

• 👀 Public preview CTA – The hero now routes “Preview productivity packs” directly to the demo and sample pack landing pages so visitors avoid the auth wall.
• ✍️ Release notes always available – MDX-based changelog entries have RSS, the hero widget, and an ISR page backed by Supabase (fallbacks to local MDX when necessary).
• 🌐 Resources SSR core – First-page rendering now ships actual content so bots/zero-JS visitors always see resources, with pagination and filters hydrating on the client.
• 🗣️ Anonymous feedback – Public NPS & comments forms (flag-gated) feed into the same storage + Sentry signals used by signed-in users.
• 🚨 Status page – Health checks expose Supabase/auth/AI status and live timestamps (linked in the footer).

Changed

• ✨ Demo & TLX – Demo results highlight Overestimation explanations with a “What this means” link and keep intent in the signup CTA.
• 🔍 SEO & i18n – Sitemap/robots exist, canonical + localized links render, and JSON-LD (Organization/Article/FAQ) ships for key pages.
• ✅ Accessibility guardrails – New axe test coverage, visible focus styles, and lazy-loaded media keep the public routes polished.

Fixed

• 🧱 Deep links preserved – Protected links now append ?next= so visitors land where they intended after auth.
• 🔐 Cron & AI – Server-only CRON_SECRET enforcement and AI proxy routes keep secrets safe.

Added

• 🔍 Open-beta conversion funnel – clarified home hero, added prominent demo CTA, and wired newsletter + feedback capture so we can learn from every visit.
• 🧪 Demo refinements – tighter unauth baseline + mini pack flow that surfaces Overestimation Index and micro‑TLX examples without requiring an account.
• 📨 Newsletter plumbing – real form, double‑opt‑in scaffolding, and analytics hooks for newsletter_submit / newsletter_confirmed.

Changed

• 🧭 Protected route UX – consistent ?next= deep‑linking across analyzer/auth flows with clear “Sign in or run demo” states instead of indefinite loaders.
• 📚 Resources IA – groundwork for server‑side pagination and trending modules so large libraries stay fast and scannable.

Fixed

• 📰 Changelog visibility – /changelog is now statically rendered from MDX, indexable, and linked from the home hero widget (no more 404 or “Loading…” only).

Added

• 🧪 3‑minute Analyzer demo – run a timed productivity pack without creating an account. Results show example Overestimation Index and TLX tiles plus a clear signup CTA to store and benchmark your data.
• 🔗 Intent-preserving auth – protected CTAs append ?next= so you land on the analyzer or resource you originally clicked after login/signup.
• 📰 MDX-based changelog + RSS – release notes now live in versioned MDX files with an RSS feed at /changelog.xml and a “What’s new” widget on the homepage.
• 🌐 SEO & structured data baseline – /robots.txt, /ai.txt, /sitemap.xml, default SEO, JSON‑LD for resources, and dynamic OG images for richer link previews.
• 📊 Admin Open Beta KPIs – new tiles for D1 activation, quiz completion rate, demo → signup conversion placeholder, high overestimation share, NPS (last 14 days), and blocked access counts.

Changed

• 🧭 Resources → Analyzer handoff – unauthenticated “Test this concept” clicks now offer a demo vs. sign‑in modal and preserve origin=resource:<id> so post‑auth flows land in the right analyzer view.
• 🧱 Landing page – added Example Metrics tiles with clearly labeled “example data” and a prominent “Try a 3‑minute demo” CTA to /analyzer/demo.

Fixed

• ♿ Accessibility + performance – added a global skip link, hardened focus states, extended jest‑axe coverage to /, /resources, /analyzer/demo, and /changelog, and ensured heavy charts are lazy‑loaded.
• ⚙️ Cron hardening & AI proxying – nightly psychometrics endpoint now enforces CRON_SECRET, and AI calls are proxied through server routes so API keys never leave the backend.