fix(quotas): unifier le décompte IA (BYOK, rollback) et combler les fuites

Centralise la réserve via ai-quota, corrige admin unavailable (-1), brancher les routes sans quota et le host-pays brainstorm, avec usage-meter élargi, noms de clusters, MCP et ajustements dashboard/insights.

Co-authored-by: Cursor <cursoragent@cursor.com>

memento-note/app/api/ai/echo/route.ts

@@ -2,6 +2,7 @@ import { NextRequest, NextResponse } from 'next/server'
 import { auth } from '@/auth'
 import { memoryEchoService } from '@/lib/ai/services/memory-echo.service'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
+import { reserveUsageOrThrow, QuotaExceededError } from '@/lib/entitlements'
 
 /**
  * GET /api/ai/echo
@@ -27,6 +28,7 @@ export async function GET(req: NextRequest) {
     }
 
     // Get next insight (respects frequency limits)
+    await reserveUsageOrThrow(session.user.id, 'reformulate')
     const insight = await memoryEchoService.getNextInsight(session.user.id)
 
     if (!insight) {
@@ -41,6 +43,9 @@ export async function GET(req: NextRequest) {
     return NextResponse.json({ insight })
 
   } catch (error) {
+    if (error instanceof QuotaExceededError) {
+      return NextResponse.json(error.toJSON(), { status: 429 })
+    }
     console.error('[/api/ai/echo] GET error:', error)
     return NextResponse.json(
       { error: 'Failed to fetch Memory Echo insight' },