fix(seo/P0): preserve human loading skeleton and crawler-readable HTML

[EagleFox31]

Human UX preserved — editorial content is identical

• Restore the original site-wide skeleton, shown both on an initial page load and during internal navigation for regular browsers.
• Suppress only the loading placeholder for known HTML-reading crawlers (Googlebot, Bingbot, OAI-SearchBot, ClaudeBot and others). This never changes an article, canonical metadata, permission check or ranking-specific content.
• Keep Next.js's built-in metadata crawler handling untouched.
• Add a zero-dependency HTTP smoke script to check full FR/EN article HTML, canonical tags, robots and sitemaps with a crawler agent.
• Unit tests verify crawler/human loading policy and that editorial data never varies by agent.

Required acceptance before merge

• GitHub CI (lint, typecheck, tests) green
• Preview: human initial load and client navigation show existing editorial skeleton
• Preview: crawler user-agent receives full HTML article with canonical metadata and without the skeleton
• Preview: article markup and content stay identical (only placeholder differs) and no loss of reader accessibility
• Preview: invalid/unpublished routes return expected noindex or 404
• Production HTTP smoke and GSC follow-up after deployment

Preview deployments have Vercel authentication protection; smoke test must be run by someone with authorized preview access before fusion. Do not merge blindly. force-dynamic and preview architecture are unchanged. Independent from Cloudflare pilot #54.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
trigenys-insight	Ready	Preview	Oct 11, 2026 10:43pm UTC