fix: admin server must use shared assets.db — stale local copy caused 'asset not found'

Root cause: Admin server's DB_PATH defaulted to its own assets.db (local copy
with 7,560 assets) instead of the main app's DB (10,148 assets). Machine ID
68026 didn't exist in the stale copy.

Fix:
- Copied main app's assets.db over (now synced)
- Set CANTEEN_DB_PATH in run.sh to default to ../canteen-asset-tracker/assets.db
  (env var can still override for custom setups)

run.sh

@@ -1,10 +1,13 @@
 #!/usr/bin/env bash
-# Run the admin server — uses same DB as main canteen-asset-tracker
-# CANTEEN_DB_PATH defaults to ../canteen-asset-tracker/assets.db
+# Run the admin server — MUST use same DB as main canteen-asset-tracker
 set -euo pipefail
 
 PORT="${PORT:-8090}"
 HOST="${HOST:-0.0.0.0}"
+APP_DIR="$(cd "$(dirname "$0")" && pwd)"
+CANTEEN_DB_PATH="${CANTEEN_DB_PATH:-${APP_DIR}/../canteen-asset-tracker/assets.db}"
+
+export CANTEEN_DB_PATH
 
 exec python -m uvicorn admin_server:app \
   --host "$HOST" \