feat: migrate semantic search to pgvector + full-text search

Replace JSON-string embeddings with native pgvector(1536) storage and
add PostgreSQL full-text search (tsvector/GIN) with Reciprocal Rank Fusion
for hybrid keyword + semantic ranking.

Changes:
- NoteEmbedding.embedding: String → vector(1536) via pgvector
- NoteEmbedding: added updatedAt for reindex tracking
- Note: added tsv (tsvector) with auto-update trigger for FTS
- semantic-search.service: hybrid FTS + vector search with RRF fusion
- embedding.service: toVectorString() for pgvector SQL literals
- Removed JS-side cosine similarity loops (now DB-side via <=>)
- Added HNSW index on NoteEmbedding.embedding (cosine distance)
- Added GIN index on Note.tsv for FTS queries

Schema migration in: prisma/migrations/20260512120000_pgvector_and_fts_search/

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

memento-note/lib/ai/tools/note-search.tool.ts

@@ -1,16 +1,16 @@
 /**
  * Note Search Tool
- * Wraps semanticSearchService.searchAsUser()
+ * Uses the unified SemanticSearchService (FTS + pgvector + RRF).
  */
 
 import { tool } from 'ai'
 import { z } from 'zod'
 import { toolRegistry } from './registry'
-import { prisma } from '@/lib/prisma'
+import { semanticSearchService } from '@/lib/ai/services/semantic-search.service'
 
 toolRegistry.register({
   name: 'note_search',
-  description: 'Search the user\'s notes using semantic search. Returns matching notes with titles and content excerpts.',
+  description: 'Search the user\'s notes using hybrid semantic + keyword search. Returns matching notes with titles and content excerpts.',
   isInternal: true,
   buildTool: (ctx) =>
     tool({
@@ -21,34 +21,20 @@ toolRegistry.register({
         notebookId: z.string().optional().describe('Optional notebook ID to restrict search to a specific notebook'),
       }),
       execute: async ({ query, limit = 5, notebookId: explicitNotebookId }) => {
-        // If no notebookId passed explicitly, fall back to the chat scope from context
         const notebookId = explicitNotebookId || ctx.notebookId
         try {
-          // Keyword fallback search using Prisma
-          const keywords = query.toLowerCase().split(/\s+/).filter(w => w.length > 2)
-          const conditions = keywords.flatMap(term => [
-            { title: { contains: term } },
-            { content: { contains: term } }
-          ])
-
-          const notes = await prisma.note.findMany({
-            where: {
-              userId: ctx.userId,
-              ...(notebookId ? { notebookId } : {}),
-              ...(conditions.length > 0 ? { OR: conditions } : {}),
-              isArchived: false,
-              trashedAt: null,
-            },
-            select: { id: true, title: true, content: true, createdAt: true },
-            take: limit,
-            orderBy: { createdAt: 'desc' },
+          const results = await semanticSearchService.searchAsUser(ctx.userId, query, {
+            limit,
+            threshold: 0.25,
+            notebookId
           })
 
-          return notes.map(n => ({
-            id: n.id,
-            title: n.title || 'Untitled',
-            excerpt: n.content.substring(0, 300),
-            createdAt: n.createdAt.toISOString(),
+          return results.map(r => ({
+            id: r.noteId,
+            title: r.title || 'Untitled',
+            excerpt: r.content.substring(0, 300),
+            score: r.score,
+            matchType: r.matchType,
           }))
         } catch (e: any) {
           return { error: `Note search failed: ${e.message}` }