Platform backbone (plan 2026-10-13-plattform-rueckgrat), all tasks reviewed:
- webhooks with delivery table, retries and stable delivery id
- calendar handlers; a finished task removes its Outlook event
- tasks, automations, inspections, varroa, treatments, feedings,
harvests, colonies, wintering, stock and bottling record domain events
in the same transaction; legacy post-commit calls removed
- row locks through one helper (lockRowForUpdate, lockRowsForShare,
lockRowsForUpdate)
- job framework with JobRun, cli exit codes, cleanup, task templates,
daily evaluation, hourly app-backup and the rekey job
- system API and page "Verwaltung > System" with retry
- @nestjs/schedule removed, guard tests against in-process schedulers
Field capture T3 (plan 2026-10-13-t3-erfassung-am-stand), all tasks reviewed:
- inspection photo offline without metadata
- bulk capture for treatments, feedings and varroa (backend + frontend)
- offline stand mode with service worker, offline pages and packs
- offline session, cold start and logout on shared devices
Docs: ARCHITECTURE.md assumptions and sections updated, TASKS.md.
Gates: backend lint, typecheck, 1770 unit, 420 e2e, build, no schema
drift; frontend lint, typecheck, 913 tests, build. Not rolled out.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Backend
- Backup covers every Prisma model (bundle v3); a guard test fails for any
new model that is neither backed up nor explicitly excluded. Unsealed
bundles may only add rows to 27 legacy master/field tables and never
overwrite; counters, ledgers, settings, roles etc. need a sealed file.
Shared import limit BACKUP_MAX_IMPORT_BYTES (default 50 MB, max 256 MB)
with warnings in log and UI; replaced document files are removed after
commit.
- Post-commit side effects (automation, calendar sync, webhooks) run via
runAfterCommit and never turn a committed write into a 500, so retries
with the same Idempotency-Key no longer create duplicates.
- Calendar outbox is multi-replica safe: per-entry claims with stale
takeover, revision-based settle, new Graph event only on 404/410,
transactionId on create, 30 s Graph timeout, URL-encoded ids, GUID
validation for groupId, "run now" in background and resets FAILED.
- Automation tasks are deduplicated under a row lock on the rule.
- Update DTOs follow one null rule: omitted = unchanged, null clears a
nullable field, null on a required field is a 400.
- Varroa mites/day is recomputed on correction; withdrawal status is
checked against the harvest date (?at=), bounded colonyIds; inspection
create is audited; updates re-enqueue calendar entries.
- Dashboard, statistics and apiary/colony guards treat WEAK, CONSPICUOUS
and QUEENLESS colonies as living.
- Prisma errors and body-parser errors are logged without query values or
request bodies.
- Migration 20261012100000_calendar_sync_claims; schema index name aligned
with the existing database object.
Frontend
- Sticky save bar sticks again (no scroll container ancestors, card uses
overflow-clip), safe-area insets, scroll-padding, FAB hidden on forms,
sticky desktop sidebar.
- Inspections record yes / no / not checked instead of silent defaults.
- All living colonies are selectable; shared RecordSelect, EnumSelect,
ChoiceGroup and TriStateField replace per-page copies; selects show
names instead of ids; labels point at real controls.
- Emptied optional fields are sent as null (emptyToNull) so every field
can be cleared; dates default to the local day and are sent via
toApiDate.
- Harvest withdrawal warning uses the harvest date; backup page warns when
a backup outgrows the import limit and lists skipped tables.
Build and docs
- Docker builds fail unless lint, typecheck and unit tests pass.
- ARCHITECTURE.md updated (assumptions 9-12, 19-21, section 8) and spec
docs/superpowers/specs/2026-10-12-plattform-rueckgrat-und-fahrplan-design.md
for the backbone (api/worker/job roles, domain event outbox) and the
feature roadmap T3-T7.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
- Deliveries connect through a lookup hook that checks the address
actually dialled, so a host that resolves to a private address after
registration (DNS rebinding) or a tampered row cannot reach internal
services. Redirects are still not followed.
- More blocked ranges (CGNAT/100.64/10 incl. cloud metadata, multicast,
reserved, benchmarking, NAT64, 6to4, site-local, IPv4-mapped IPv6);
only ports 80/443, no credentials in the URL, https in production.
- 5 s timeout, response body discarded; dispatch no longer waits for
receivers, so a hanging endpoint cannot stall the user's request.
- Signature covers "timestamp.deliveryId.body" with X-ApiCore-Timestamp
and X-ApiCore-Delivery headers, so receivers can reject replays.
- Signing secrets are stored AES-256-GCM-encrypted; older plain rows still
work.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Letzte Phase des RBAC-Umbaus: die transitionale Rollen-Infrastruktur,
die seit Phase 1 parallel zum neuen Role/Permission-System mitlief,
ist jetzt komplett weg.
- Prisma-Migration: User.legacyRole-Spalte und LegacyRole-Enum
gedroppt (nichts liest die Spalte mehr, seit Phase 2 jeder Controller
auf @RequirePermission laeuft).
- roles.guard.ts, roles.guard.spec.ts, roles.decorator.ts geloescht;
RolesGuard aus den APP_GUARD-Providern entfernt.
- AuthenticatedUser/JWT-Payload verlieren das role-Feld (permissions
werden ohnehin frisch pro Request aus der DB aufgeloest, nicht aus
dem Claim).
- Users-API: CreateUserDto/UpdateUserDto akzeptieren jetzt roleId
(eine echte Role.id) statt der alten role-Enum-Werte; UsersService
validiert, dass die Rolle existiert und aktiv ist, bevor sie
zugewiesen wird.
- Alle 34 e2e-Fixtures und der Seed-Script folgen dem neuen Schema.
Verifiziert: tsc, lint, build, unit (220/220), volle e2e-Suite
159/159 gruen - keine uebersprungenen Tests.
Co-Authored-By: Claude Sonnet 5 <[email protected]>
Replaces the fixed 5-value Role enum with a database-driven Role/Permission
model: admins can create custom roles with a precise per-resource/action
permission set (83-key code-owned catalog), instead of being locked to
ADMIN/BEEKEEPER/STAFF/READONLY/API_CLIENT.
- New Role/Permission/RolePermission tables; legacy Role enum renamed to
LegacyRole and kept as User.legacyRole so all existing @Roles() call
sites keep working unchanged during the migration.
- New RolesController/RolesService (CRUD, permission-set editing,
deactivate/reactivate) with lockout-prevention invariants: system roles
(ADMIN) can't be edited or deactivated, and a role still assigned to
users can't be deactivated.
- New PermissionResolverService (30s-cached, DB-fresh per user) and
PermissionsGuard + @RequirePermission(), registered as an additional
APP_GUARD alongside the legacy RolesGuard so both systems run
side-by-side until controllers are migrated in Phase 2.
- PermissionCatalogSyncService seeds any new catalog entries into the DB
on boot, so future features only touch the TS catalog.
- New GET /auth/me returns the caller's role and effective permissions.
- Migration hand-written (destructive column rename) with grants derived
from a fresh grep of every existing @Roles() call site, verified
against 5 controllers read in full for STAFF's exact write boundary.
- test/roles.e2e-spec.ts covers the new endpoints and invariants end to
end; test/authorization-coverage.e2e-spec.ts added but skipped until
Phase 2 gives every controller a @RequirePermission.
Co-Authored-By: Claude Sonnet 5 <[email protected]>
WebhooksService.create() rejected obviously private, loopback, or
link-local URLs (incl. cloud metadata 169.254.169.254 and the
decimal/octal/hex IPv4 shorthand encodings the WHATWG URL parser
normalizes) so an admin-configured webhook can no longer be used to
proxy requests into internal infrastructure. Registration-time only,
no DNS resolution (DNS-rebinding is out of scope for this check).
WebhooksModule (WebhookSubscription: URL, Events, Secret, aktiv) mit
Event-Dispatcher, angehaengt an 10 Ereignisse in 8 bestehenden Services
(colony.created, inspection.created, queen.created, queen.replaced,
harvest.created, batch.created, batch.updated, product.stock_changed,
task.created, task.completed). Auslieferung ist best-effort und blockiert
die ausloesende Anfrage nicht bei Fehlern.
