feat: grouping phase 1 — schema, ungrouped tab, time-window grouping, hash verification

Schema:
- Add GroupingSource enum (ALBUM, MANUAL, AUTO_TIME, AUTO_PATTERN, etc.)
- Add groupingSource field to PackageGroup with backfill
- Add SystemNotification model for persistent alerts
- Add NotificationType and NotificationSeverity enums

Ungrouped staging tab:
- Add listUngroupedPackages/countUngroupedPackages queries
- Add "Ungrouped" tab to STL page showing packages without a group

Time-window auto-grouping:
- After album grouping, cluster ungrouped packages within configurable
  time window (default 5 min, AUTO_GROUP_TIME_WINDOW_MINUTES env var)
- Groups named from common filename prefix
- Groups created with groupingSource=AUTO_TIME

Hash verification after split:
- Re-hash split parts and compare to original contentHash
- Log error and create SystemNotification on mismatch
- Prevents silently corrupted split uploads

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

worker/src/worker.ts

@@ -47,7 +47,8 @@ import { readRarContents } from "./archive/rar-reader.js";
 import { read7zContents } from "./archive/sevenz-reader.js";
 import { byteLevelSplit, concatenateFiles } from "./archive/split.js";
 import { uploadToChannel } from "./upload/channel.js";
-import { processAlbumGroups, type IndexedPackageRef } from "./grouping.js";
+import { processAlbumGroups, processTimeWindowGroups, type IndexedPackageRef } from "./grouping.js";
+import { db } from "./db/client.js";
 import type { TelegramAccount, TelegramChannel } from "@prisma/client";
 import type { Client } from "tdl";
 
@@ -790,6 +791,9 @@ async function processArchiveSets(
       indexedPackageRefs,
       scanResult.photos
     );
+
+    // Time-window grouping for remaining ungrouped packages
+    await processTimeWindowGroups(channel.id, indexedPackageRefs);
   }
 
   return maxProcessedId;
@@ -1053,6 +1057,43 @@ async function processOneArchiveSet(
       uploadPaths = splitPaths;
     }
 
+    // ── Hash verification after split ──
+    // If we split/repacked, verify the split parts hash matches the original
+    if (splitPaths.length > 0) {
+      const splitHash = await hashParts(splitPaths);
+      if (splitHash !== contentHash) {
+        accountLog.error(
+          { fileName: archiveName, originalHash: contentHash, splitHash, parts: splitPaths.length },
+          "Hash mismatch after split — file may be corrupted"
+        );
+        // Record notification for visibility
+        try {
+          await db.systemNotification.create({
+            data: {
+              type: "HASH_MISMATCH",
+              severity: "ERROR",
+              title: `Hash mismatch after splitting ${archiveName}`,
+              message: `Expected ${contentHash.slice(0, 16)}… but got ${splitHash.slice(0, 16)}… after splitting into ${splitPaths.length} parts`,
+              context: {
+                fileName: archiveName,
+                originalHash: contentHash,
+                splitHash,
+                partCount: splitPaths.length,
+                sourceChannelId: channel.id,
+              },
+            },
+          });
+        } catch {
+          // Best-effort notification
+        }
+        throw new Error(`Hash mismatch after split for ${archiveName}: expected ${contentHash}, got ${splitHash}`);
+      }
+      accountLog.debug(
+        { fileName: archiveName, hash: contentHash.slice(0, 16), parts: splitPaths.length },
+        "Split hash verified — matches original"
+      );
+    }
+
     // ── Uploading ──
     // Check if a prior run already uploaded this file (orphaned upload scenario:
     // file reached Telegram but DB write failed or worker crashed before indexing)