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/auto-labels/route.ts

@@ -4,6 +4,7 @@ import { auth } from '@/auth'
 import { autoLabelCreationService } from '@/lib/ai/services'
 import { getAISettings } from '@/app/actions/ai-settings'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
+import { withAiQuota, handleQuotaHttpError } from '@/lib/ai-quota'
 
 /**
  * POST /api/ai/auto-labels - Suggest new labels for a notebook
@@ -59,11 +60,16 @@ export async function POST(request: NextRequest) {
       )
     }
 
-    // Get label suggestions
-    const suggestions = await autoLabelCreationService.suggestLabels(
-      notebookId,
+    const suggestions = await withAiQuota(
       session.user.id,
-      language
+      'auto_tag',
+      () =>
+        autoLabelCreationService.suggestLabels(
+          notebookId,
+          session.user!.id,
+          language,
+        ),
+      { lane: 'tags' },
     )
 
     if (!suggestions) {
@@ -79,8 +85,10 @@ export async function POST(request: NextRequest) {
       data: suggestions,
     })
   } catch (error) {
+    const quotaResp = handleQuotaHttpError(error)
+    if (quotaResp) return quotaResp
     console.error('[/api/ai/auto-labels] POST failed:', error)
-    return NextResponse.json({ success: true, data: null })
+    return NextResponse.json({ success: false, error: 'auto_labels_failed' }, { status: 500 })
   }
 }
 

memento-note/app/api/ai/batch-organize/route.ts

@@ -2,6 +2,7 @@ import { NextRequest, NextResponse } from 'next/server'
 import { auth } from '@/auth'
 import { batchOrganizationService } from '@/lib/ai/services'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
+import { reserveUsageOrThrow, QuotaExceededError } from '@/lib/entitlements'
 
 /**
  * POST /api/ai/batch-organize - Create organization plan for notes in Inbox
@@ -40,6 +41,8 @@ export async function POST(request: NextRequest) {
       }
     }
 
+    await reserveUsageOrThrow(session.user.id, 'reformulate')
+
     // Create organization plan
     const plan = await batchOrganizationService.createOrganizationPlan(
       session.user.id,

memento-note/app/api/ai/describe-image/route.ts

@@ -3,6 +3,7 @@ import { auth } from '@/auth'
 import { getAISettings } from '@/app/actions/ai-settings'
 import { describeImages } from '@/lib/ai/services/image-description.service'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
+import { withAiQuota, handleQuotaHttpError } from '@/lib/ai-quota'
 
 export async function POST(req: NextRequest) {
   try {
@@ -11,7 +12,6 @@ export async function POST(req: NextRequest) {
       return NextResponse.json({ error: 'Unauthorized' }, { status: 401 })
     }
 
-    // GDPR AI Consent check
     if (!(await hasUserAiConsent())) {
       return NextResponse.json({ error: 'ai_consent_required' }, { status: 403 })
     }
@@ -27,23 +27,31 @@ export async function POST(req: NextRequest) {
       return NextResponse.json({ error: 'imageUrls must be a non-empty array' }, { status: 400 })
     }
 
-    const result = await describeImages(
-      imageUrls,
-      mode === 'title' ? 'title' : 'description',
-      language || 'fr'
+    const isTitleMode = mode === 'title'
+    const feature = isTitleMode ? 'auto_title' : 'reformulate'
+    const lane = isTitleMode ? 'tags' : 'chat'
+
+    const result = await withAiQuota(
+      session.user.id,
+      feature,
+      () => describeImages(
+        imageUrls,
+        isTitleMode ? 'title' : 'description',
+        language || 'fr'
+      ),
+      { lane },
     )
 
-    // For title mode, return suggestions in same format as /api/ai/title-suggestions
-    if (mode === 'title' && result.suggestions) {
+    if (isTitleMode && result.suggestions) {
       return NextResponse.json({ suggestions: result.suggestions })
     }
 
     return NextResponse.json(result)
-  } catch (error: any) {
+  } catch (error: unknown) {
+    const quotaResp = handleQuotaHttpError(error)
+    if (quotaResp) return quotaResp
     console.error('[describe-image] Error:', error)
-    return NextResponse.json(
-      { error: error.message || 'Failed to describe image' },
-      { status: 500 }
-    )
+    const message = error instanceof Error ? error.message : 'Failed to describe image'
+    return NextResponse.json({ error: message }, { status: 500 })
   }
 }

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

@@ -4,6 +4,7 @@ import { getChatProvider } from '@/lib/ai/factory'
 import { getSystemConfig } from '@/lib/config'
 import prisma from '@/lib/prisma'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
+import { withAiQuota, handleQuotaHttpError } from '@/lib/ai-quota'
 
 /**
  * POST /api/ai/echo/fusion
@@ -34,7 +35,6 @@ export async function POST(req: NextRequest) {
       )
     }
 
-    // Fetch the notes
     const notes = await prisma.note.findMany({
       where: {
         id: { in: noteIds },
@@ -55,11 +55,9 @@ export async function POST(req: NextRequest) {
       )
     }
 
-    // Get AI provider
     const config = await getSystemConfig()
     const provider = getChatProvider(config)
 
-    // Build fusion prompt
     const notesDescriptions = notes.map((note, index) => {
       return `Note ${index + 1}: "${note.title || 'Untitled'}"
 ${note.content}`
@@ -90,21 +88,21 @@ Output format:
 
 Begin:`
 
-    try {
-      const fusedContent = await provider.generateText(fusionPrompt)
-
-      return NextResponse.json({
-        fusedNote: fusedContent,
-        notesCount: notes.length
-      })
-    } catch (error) {
-      return NextResponse.json(
-        { error: 'Failed to generate fusion' },
-        { status: 500 }
-      )
-    }
+    const fusedContent = await withAiQuota(
+      session.user.id,
+      'reformulate',
+      () => provider.generateText(fusionPrompt),
+      { lane: 'chat' },
+    )
 
+    return NextResponse.json({
+      fusedNote: fusedContent,
+      notesCount: notes.length
+    })
   } catch (error) {
+    const quotaResp = handleQuotaHttpError(error)
+    if (quotaResp) return quotaResp
+    console.error('[/api/ai/echo/fusion] Error:', error)
     return NextResponse.json(
       { error: 'Failed to process fusion request' },
       { status: 500 }

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' },

memento-note/app/api/ai/enrich-from-resource/route.ts

@@ -3,6 +3,7 @@ import { auth } from '@/auth'
 import { getSystemConfig } from '@/lib/config'
 import { getTagsProvider } from '@/lib/ai/factory'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
+import { withAiQuota, handleQuotaHttpError } from '@/lib/ai-quota'
 
 export async function POST(request: NextRequest) {
   try {
@@ -32,7 +33,6 @@ export async function POST(request: NextRequest) {
     let prompt: string
 
     if (mode === 'complete') {
-      // Add missing info from resource without rewriting the existing note
       prompt = `You are an expert note editor. Your task is to enrich an existing note by adding relevant information from a provided resource, WITHOUT modifying or rewriting the existing content.
 
 LANGUAGE RULE: Respond in ${lang}. Match the language of the existing note.
@@ -56,7 +56,6 @@ INSTRUCTIONS:
 - Format the new content consistently with the existing note style and the requested FORMAT RULE
 - Respond ONLY with the enriched note content, no explanations`
     } else {
-      // Merge: intelligently rewrite integrating both sources
       prompt = `You are an expert note writer. Your task is to intelligently merge an existing note with a resource into a single, coherent, well-structured document.
 
 LANGUAGE RULE: Respond in ${lang}. Match the language of the existing note.
@@ -81,14 +80,19 @@ INSTRUCTIONS:
 - Respond ONLY with the merged content, no meta-commentary or explanations`
     }
 
-    const enrichedContent = await provider.generateText(prompt)
+    const enrichedContent = await withAiQuota(
+      session.user.id,
+      'reformulate',
+      () => provider.generateText(prompt),
+      { lane: 'chat' },
+    )
 
     return NextResponse.json({ enrichedContent: enrichedContent.trim() })
-  } catch (error: any) {
+  } catch (error: unknown) {
+    const quotaResp = handleQuotaHttpError(error)
+    if (quotaResp) return quotaResp
     console.error('[enrich-from-resource] Error:', error)
-    return NextResponse.json(
-      { error: error.message || 'Failed to enrich content' },
-      { status: 500 }
-    )
+    const message = error instanceof Error ? error.message : 'Failed to enrich content'
+    return NextResponse.json({ error: message }, { status: 500 })
   }
 }

memento-note/app/api/ai/notebook-summary/route.ts

@@ -3,6 +3,7 @@ import { NextRequest, NextResponse } from 'next/server'
 import { auth } from '@/auth'
 import { notebookSummaryService } from '@/lib/ai/services'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
+import { reserveUsageOrThrow, QuotaExceededError } from '@/lib/entitlements'
 
 /**
  * POST /api/ai/notebook-summary - Generate summary for a notebook
@@ -51,6 +52,8 @@ export async function POST(request: NextRequest) {
       )
     }
 
+    await reserveUsageOrThrow(session.user.id, 'reformulate')
+
     // Generate summary
     const summary = await notebookSummaryService.generateSummary(
       notebookId,

memento-note/app/api/ai/organize-notebook/route.ts

@@ -2,7 +2,7 @@ import { NextRequest, NextResponse } from 'next/server'
 import { auth } from '@/auth'
 import prisma from '@/lib/prisma'
 import { notebookOrganizerService } from '@/lib/ai/services/notebook-organizer.service'
-import { reserveUsageOrThrow, QuotaExceededError } from '@/lib/entitlements'
+import { withAiQuota, handleQuotaHttpError } from '@/lib/ai-quota'
 import { syncNoteLabels } from '@/app/actions/notes'
 
 export async function POST(request: NextRequest) {
@@ -17,19 +17,6 @@ export async function POST(request: NextRequest) {
       return NextResponse.json({ error: 'notebookId is required' }, { status: 400 })
     }
 
-    try {
-      await reserveUsageOrThrow(session.user.id, 'reformulate')
-    } catch (err) {
-      if (err instanceof QuotaExceededError) {
-        const isTierLocked = err.currentQuota === 0
-        return NextResponse.json(
-          { error: isTierLocked ? 'feature_locked' : 'quota_exceeded', errorKey: isTierLocked ? 'ai.featureLocked' : 'ai.quotaExceeded' },
-          { status: 402 },
-        )
-      }
-      throw err
-    }
-
     const notes = await prisma.note.findMany({
       where: { notebookId, trashedAt: null, userId: session.user.id },
       select: { id: true, title: true, content: true },
@@ -46,13 +33,20 @@ export async function POST(request: NextRequest) {
       contentPreview: n.content.replace(/<[^>]+>/g, ' ').replace(/\s+/g, ' ').trim().slice(0, 300),
     }))
 
-    const result = await notebookOrganizerService.analyze(notesForAnalysis)
-
+    const result = await withAiQuota(
+      session.user.id,
+      'reformulate',
+      () => notebookOrganizerService.analyze(notesForAnalysis),
+      { lane: 'chat' },
+    )
 
     return NextResponse.json(result)
-  } catch (error: any) {
+  } catch (error: unknown) {
+    const quotaResp = handleQuotaHttpError(error)
+    if (quotaResp) return quotaResp
     console.error('[Notebook Organizer] Error:', error)
-    return NextResponse.json({ error: error.message || 'Failed to organize notebook' }, { status: 500 })
+    const message = error instanceof Error ? error.message : 'Failed to organize notebook'
+    return NextResponse.json({ error: message }, { status: 500 })
   }
 }
 

memento-note/app/api/ai/search-overview/route.ts

@@ -1,7 +1,7 @@
-import { rateLimit } from '@/lib/rate-limit'
 import { NextRequest, NextResponse } from 'next/server'
 import { auth } from '@/auth'
 import { searchOverviewService } from '@/lib/ai/services/search-overview.service'
+import { withAiQuota, handleQuotaHttpError } from '@/lib/ai-quota'
 
 export async function POST(request: NextRequest) {
   try {
@@ -15,11 +15,19 @@ export async function POST(request: NextRequest) {
       return NextResponse.json({ error: 'query and results required' }, { status: 400 })
     }
 
-    const overview = await searchOverviewService.generate(query, results, language)
+    const overview = await withAiQuota(
+      session.user.id,
+      'semantic_search',
+      () => searchOverviewService.generate(query, results, language),
+      { lane: 'chat' },
+    )
 
     return NextResponse.json(overview)
-  } catch (error: any) {
+  } catch (error: unknown) {
+    const quotaResp = handleQuotaHttpError(error)
+    if (quotaResp) return quotaResp
     console.error('[Search Overview] Error:', error)
-    return NextResponse.json({ error: error.message, hasRelevantInfo: false, answer: '' }, { status: 500 })
+    const message = error instanceof Error ? error.message : 'Search overview failed'
+    return NextResponse.json({ error: message, hasRelevantInfo: false, answer: '' }, { status: 500 })
   }
 }

memento-note/app/api/ai/suggest-notebook/route.ts

@@ -2,6 +2,7 @@ import { NextRequest, NextResponse } from 'next/server'
 import { auth } from '@/auth'
 import { notebookSuggestionService } from '@/lib/ai/services/notebook-suggestion.service'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
+import { withAiQuota, handleQuotaHttpError } from '@/lib/ai-quota'
 
 export async function POST(req: NextRequest) {
   try {
@@ -21,7 +22,6 @@ export async function POST(req: NextRequest) {
       return NextResponse.json({ error: 'noteContent is required' }, { status: 400 })
     }
 
-    // Minimum content length for suggestion (20 words as per specs)
     const wordCount = noteContent.trim().split(/\s+/).length
     if (wordCount < 20) {
       return NextResponse.json({
@@ -31,18 +31,25 @@ export async function POST(req: NextRequest) {
       })
     }
 
-    // Get suggestion from AI service
-    const suggestedNotebook = await notebookSuggestionService.suggestNotebook(
-      noteContent,
+    const suggestedNotebook = await withAiQuota(
       session.user.id,
-      language
+      'auto_tag',
+      () => notebookSuggestionService.suggestNotebook(
+        noteContent,
+        session.user!.id,
+        language
+      ),
+      { lane: 'tags' },
     )
 
     return NextResponse.json({
       suggestion: suggestedNotebook,
-      confidence: suggestedNotebook ? 0.8 : 0 // Placeholder confidence score
+      confidence: suggestedNotebook ? 0.8 : 0
     })
   } catch (error) {
+    const quotaResp = handleQuotaHttpError(error)
+    if (quotaResp) return quotaResp
+    console.error('[/api/ai/suggest-notebook] Error:', error)
     return NextResponse.json(
       { error: 'Failed to generate suggestion' },
       { status: 500 }

memento-note/app/api/ai/title-suggestions/route.ts

@@ -1,10 +1,11 @@
 import { rateLimit } from '@/lib/rate-limit'
 import { NextRequest, NextResponse } from 'next/server'
-import { runLaneWithBillingUser, willUseByokForLane } from '@/lib/ai/provider-for-user'
+import { runLaneWithBillingUser } from '@/lib/ai/provider-for-user'
 import { getSystemConfig } from '@/lib/config'
 import { auth } from '@/auth'
 import { getAISettings } from '@/app/actions/ai-settings'
-import { reserveUsageOrThrow, QuotaExceededError, QuotaServiceUnavailableError } from '@/lib/entitlements'
+import { reserveAiUsageOrThrow } from '@/lib/ai-quota'
+import { QuotaExceededError, QuotaServiceUnavailableError } from '@/lib/entitlements'
 import { z } from 'zod'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
 
@@ -61,19 +62,16 @@ export async function POST(req: NextRequest) {
     }
 
     const config = await getSystemConfig()
-    const { usedByok: willUseByok } = await willUseByokForLane('tags', config, session.user.id)
-    if (!willUseByok) {
-      try {
-        await reserveUsageOrThrow(session.user.id, 'auto_title')
-      } catch (err) {
-        if (err instanceof QuotaExceededError) {
-          return NextResponse.json(err.toJSON(), { status: 402 })
-        }
-        if (err instanceof QuotaServiceUnavailableError || process.env.NODE_ENV === 'production') {
-          return NextResponse.json({ error: 'QUOTA_SERVICE_UNAVAILABLE' }, { status: 503 })
-        }
-        console.error('[/api/ai/title-suggestions] Quota check error (fail-open):', err)
+    try {
+      await reserveAiUsageOrThrow(session.user.id, 'auto_title', { lane: 'tags' })
+    } catch (err) {
+      if (err instanceof QuotaExceededError) {
+        return NextResponse.json(err.toJSON(), { status: 402 })
       }
+      if (err instanceof QuotaServiceUnavailableError || process.env.NODE_ENV === 'production') {
+        return NextResponse.json({ error: 'QUOTA_SERVICE_UNAVAILABLE' }, { status: 503 })
+      }
+      console.error('[/api/ai/title-suggestions] Quota check error (fail-open):', err)
     }
 
     // Détecter la langue du contenu (simple détection basée sur les caractères et mots)

memento-note/app/api/ai/transform-markdown/route.ts

@@ -3,6 +3,7 @@ import { auth } from '@/auth'
 import { getChatProvider } from '@/lib/ai/factory'
 import { getSystemConfig } from '@/lib/config'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
+import { withAiQuota, handleQuotaHttpError } from '@/lib/ai-quota'
 
 export async function POST(request: NextRequest) {
   try {
@@ -18,12 +19,10 @@ export async function POST(request: NextRequest) {
 
     const { text } = await request.json()
 
-    // Validation
     if (!text || typeof text !== 'string') {
       return NextResponse.json({ error: 'Text is required' }, { status: 400 })
     }
 
-    // Validate word count
     const wordCount = text.split(/\s+/).length
     if (wordCount < 10) {
       return NextResponse.json(
@@ -42,11 +41,9 @@ export async function POST(request: NextRequest) {
     const config = await getSystemConfig()
     const provider = getChatProvider(config)
 
-    // Detect language from text
     const hasFrench = /[àâäéèêëïîôùûüÿç]/i.test(text)
     const responseLanguage = hasFrench ? 'French' : 'English'
 
-    // Build prompt to transform text to Markdown
     const prompt = hasFrench
       ? `Tu es un expert en Markdown. Transforme ce texte ${responseLanguage} en Markdown bien formaté.
 
@@ -77,19 +74,22 @@ ${text}
 
 Respond ONLY with the transformed Markdown text, no explanations.`
 
-    
-    const transformedText = await provider.generateText(prompt)
+    const transformedText = await withAiQuota(
+      session.user.id,
+      'reformulate',
+      () => provider.generateText(prompt),
+      { lane: 'chat' },
+    )
 
-    
     return NextResponse.json({
       originalText: text,
-      transformedText: transformedText,
+      transformedText,
       language: responseLanguage
     })
-  } catch (error: any) {
-    return NextResponse.json(
-      { error: error.message || 'Failed to transform text to Markdown' },
-      { status: 500 }
-    )
+  } catch (error: unknown) {
+    const quotaResp = handleQuotaHttpError(error)
+    if (quotaResp) return quotaResp
+    const message = error instanceof Error ? error.message : 'Failed to transform text to Markdown'
+    return NextResponse.json({ error: message }, { status: 500 })
   }
 }

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

@@ -3,6 +3,7 @@ import { auth } from '@/auth'
 import { getTagsProvider } from '@/lib/ai/factory'
 import { getSystemConfig } from '@/lib/config'
 import { hasUserAiConsent } from '@/lib/consent/server-consent'
+import { reserveUsageOrThrow, QuotaExceededError } from '@/lib/entitlements'
 
 export async function POST(request: NextRequest) {
   try {
@@ -24,12 +25,18 @@ export async function POST(request: NextRequest) {
     const config = await getSystemConfig()
     const provider = getTagsProvider(config)
 
+    await reserveUsageOrThrow(session.user.id, 'reformulate')
+
     const prompt = `Translate the following text to ${targetLanguage}. Return ONLY the translated text, no explanation, no preamble, no quotes:\n\n${text}`
 
     const translatedText = await provider.generateText(prompt)
 
     return NextResponse.json({ translatedText: translatedText.trim() })
-  } catch (error: any) {
-    return NextResponse.json({ error: error.message || 'Translation failed' }, { status: 500 })
+  } catch (error: unknown) {
+    if (error instanceof QuotaExceededError) {
+      return NextResponse.json(error.toJSON(), { status: 429 })
+    }
+    const message = error instanceof Error ? error.message : 'Translation failed'
+    return NextResponse.json({ error: message }, { status: 500 })
   }
 }