🛠️ Phase 7: Visual & IA Overhaul (in progress)

Triggered by a UX audit (browser-driven) + Saif's reaction: the feature-complete English site read as "boring / cheap / off", text-heavy, no imagery/icons/badges, dead scroll space, empty-looking graph/analytics, single-page with no nav, and, worst, live N stats + [TODO] placeholders in case-study Outcomes. Audit scored it 6/10 ("smart skeleton shipped a week early"; taste + engineering senior, finish lacking).

Scope (Fable prompt FABLE-PHASE-7-OVERHAUL.md, parts A–H):

• A: multi-page + persistent header nav + page transitions; kill dead space (D018 - Multi-page and Header Nav).
• B: remove all visible N / [TODO] placeholders; fix Prism copy inconsistency.
• C: visual richness, real web screenshots, device mockups + store badges for apps, tech icons, logos (D019 - Visual Richness Overhaul).
• D: rebuild Work section (thumbnails, balanced layout, fix hover scroll-jump).
• E: case studies → scrollytelling + next-project nav.
• F: fix graph (typed/iconned nodes + legend, readable labels, reliable drag, click-to-focus) + blank homepage teaser canvas.
• G: analytics + loading + AI-error states look designed.
• H: motion polish + RTL bidi fix (<bdi> for English runs/numbers in ar/hi).

Keep: typography, Stone palette + one restrained accent (no rainbow), light/dark, i18n/RTL, analytics/graph/AI engineering, perf (Lighthouse 95+), axe AA, deferred-service states (D017), English-only.

Blocked on Saif for full quality: real homepage stat integers, real case-study Outcome metrics (incl. safe Shoppin' numbers), real app screenshots. Prompt degrades gracefully (device-mockup placeholders + qualitative copy) so nothing broken ships meanwhile.

Then: Phase 8, launch.