Deduplicate performance snapshots: one row per platform per day

Added snapshot_date column with unique constraint on (platform_id,
snapshot_date). Worker now upserts instead of inserting, so repeated
scrapes on the same day update the existing row rather than creating
duplicates that cause inflated chart values. Cleaned up existing
duplicate rows in migration.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

apps/api/src/db/schema.ts

@@ -42,6 +42,7 @@ export const performanceSnapshots = pgTable('performance_snapshots', {
     .notNull()
     .references(() => platforms.id),
   capturedAt: timestamptz('captured_at').defaultNow().notNull(),
+  snapshotDate: date('snapshot_date').defaultNow().notNull(),
   periodLabel: varchar('period_label').notNull(),
   viewsSearch: integer('views_search'),
   viewsListing: integer('views_listing'),
@@ -51,7 +52,9 @@ export const performanceSnapshots = pgTable('performance_snapshots', {
   avgDailyRate: numeric('avg_daily_rate'),
   revenueTotal: numeric('revenue_total'),
   rawJson: jsonb('raw_json'),
-});
+}, (table) => [
+  unique('perf_snap_platform_date_uniq').on(table.platformId, table.snapshotDate),
+]);
 
 export const performanceSnapshotsRelations = relations(
   performanceSnapshots,

apps/api/src/routes/performance.ts

@@ -122,7 +122,7 @@ export default async function (fastify: FastifyInstance) {
 
       const rows = await db
         .select({
-          date: performanceSnapshots.periodLabel,
+          date: performanceSnapshots.snapshotDate,
           capturedAt: performanceSnapshots.capturedAt,
           platformId: performanceSnapshots.platformId,
           viewsSearch: performanceSnapshots.viewsSearch,