Seed batches: test chunked inserts with a lowered chunk size

The 6,000-file test took 29 s and failed CI; 120 files in chunks of 50
cover the same path.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

app/services/SeedService.ts

@@ -1,6 +1,6 @@
 import db from '@adonisjs/lucid/services/db'
 import IngestService from '#services/IngestService'
-import FileStore, { INSERT_ROWS } from '#services/FileStore'
+import FileStore, { insertLimits } from '#services/FileStore'
 import SfvService from '#services/SfvService'
 import ContentService from '#services/ContentService'
 import QueueService from '#services/QueueService'
@@ -78,9 +78,9 @@ class SeedService {
       file_id: fileIds.get(file.fileKey) ?? null,
     }))
     binaries.sort((a, b) => a.file_key.localeCompare(b.file_key))
-    for (let i = 0; i < binaries.length; i += INSERT_ROWS) {
+    for (let i = 0; i < binaries.length; i += insertLimits.rows) {
       await knex('binaries')
-        .insert(binaries.slice(i, i + INSERT_ROWS))
+        .insert(binaries.slice(i, i + insertLimits.rows))
         .onConflict('file_key')
         .merge({ file_id: knex.raw('coalesce(binaries.file_id, excluded.file_id)') as any })
     }