Closes a bypass surfaced by the 2026-04 audit probe (axis-1 Q2): any
authenticated user could POST /api/v1/subscriptions/subscribe on a paid
plan and receive 201 active without the payment provider ever being
invoked. The resulting row satisfied `checkEligibility()` in the
distribution service via `can_sell_on_marketplace=true` on the Creator
plan — effectively free access to /api/v1/distribution/submit, which
dispatches to external partners.
Fix is centralised in `GetUserSubscription` so there is no code path
that can grant subscription-gated access without routing through the
payment check. Effective-payment = free plan OR unexpired trial OR
invoice with non-empty hyperswitch_payment_id. Migration 980 sweeps
pre-existing fantôme rows into `expired`, preserving the tuple in a
dated audit table for support outreach.
Subscribe and subscribeToFreePlan treat the new ErrSubscriptionNoPayment
as equivalent to ErrNoActiveSubscription so re-subscription works
cleanly post-cleanup. GET /me/subscription surfaces needs_payment=true
with a support-contact message rather than a misleading "you're on
free" or an opaque 500. TODO(v1.0.7-item-G) annotation marks where the
`if s.paymentProvider != nil` short-circuit needs to become a mandatory
pending_payment state.
Probe script `scripts/probes/subscription-unpaid-activation.sh` kept as
a versioned regression test — dry-run by default, --destructive logs in
and attempts the exploit against a live backend with automatic cleanup.
8-case unit test matrix covers the full hasEffectivePayment predicate.
Smoke validated end-to-end against local v1.0.6.2: POST /subscribe
returns 201 (by design — item G closes the creation path), but
GET /me/subscription returns subscription=null + needs_payment=true,
distribution eligibility returns false.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- ORDER BY dynamiques : whitelist explicite, fallback created_at DESC
- Login/register soumis au rate limiter global
- VERSION sync + check CI
- Nettoyage références veza-chat-server
- Go 1.24 partout (Dockerfile, workflows)
- TODO/FIXME/HACK convertis en issues ou résolus
- Completed Action 1.3.1.2: Tested 36 endpoints for response format consistency
- Fixed test script to handle subshell issues with RESULTS array
- Created ENDPOINT_FORMAT_AUDIT.md documenting findings
- Found 2 endpoints using wrapped format, 0 direct format
- Most endpoints require auth (22) or have errors (12)
- Limited coverage due to authentication requirements and path parameters
- Completed Action 1.3.1.1: Created test-endpoint-formats.sh
- Script reads endpoints from Swagger spec and tests each one
- Identifies wrapped vs direct response formats
- Outputs JSON report with format categorization
- Handles auth-required endpoints gracefully
- Can be run against any base URL
- Configure LOG_DIR=/var/log/veza pour tous les services
- Ajoute scripts de gestion des logs (setup, view, rotate)
- Configure volume Docker partagé pour les logs
- Logs organisés par service avec fichiers séparés pour les erreurs
- Rotation automatique : 100MB, 10 backups, 30 jours, compression gzip
- Documentation dans LOGGING.md et ENV_CONFIG.md
Services configurés:
- Backend API: backend-api.log, redis.log, db.log, rabbitmq.log
- Chat Server: chat-server.log (à configurer)
- Stream Server: stream-server.log (à configurer)
Le backend API a déjà toute l'infrastructure de logging en place.
Les serveurs chat et stream utiliseront LOG_DIR depuis l'environnement.
- Added TokenVersion: 0 to user creation in Register service
- This field is required (NOT NULL) in the database
- Backend needs to be restarted for this fix to take effect
- Stop execution if register fails (don't try login with non-existent user)
- Add warning when register fails (backend may need restart)
- Skip login test if register failed
- Better error messages
- Updated to extract from .data.token.access_token (correct format)
- Added fallback patterns for different response formats
- Added debug logging when token extraction fails
- Fixed refresh token extraction as well
- Changed password_confirmation to password_confirm in test-mvp-api.sh
- Format now matches backend DTO (password_confirm)
- Register still fails with code 9000 (DB/validation issue - BUG-004)
- Updated MVP_BUGS_TODOLIST.json with progress
Backend Go:
- Remplacement complet des anciennes migrations par la base V1 alignée sur ORIGIN.
- Durcissement global du parsing JSON (BindAndValidateJSON + RespondWithAppError).
- Sécurisation de config.go, CORS, statuts de santé et monitoring.
- Implémentation des transactions P0 (RBAC, duplication de playlists, social toggles).
- Ajout d’un job worker structuré (emails, analytics, thumbnails) + tests associés.
- Nouvelle doc backend : AUDIT_CONFIG, BACKEND_CONFIG, AUTH_PASSWORD_RESET, JOB_WORKER_*.
Chat server (Rust):
- Refonte du pipeline JWT + sécurité, audit et rate limiting avancé.
- Implémentation complète du cycle de message (read receipts, delivered, edit/delete, typing).
- Nettoyage des panics, gestion d’erreurs robuste, logs structurés.
- Migrations chat alignées sur le schéma UUID et nouvelles features.
Stream server (Rust):
- Refonte du moteur de streaming (encoding pipeline + HLS) et des modules core.
- Transactions P0 pour les jobs et segments, garanties d’atomicité.
- Documentation détaillée de la pipeline (AUDIT_STREAM_*, DESIGN_STREAM_PIPELINE, TRANSACTIONS_P0_IMPLEMENTATION).
Documentation & audits:
- TRIAGE.md et AUDIT_STABILITY.md à jour avec l’état réel des 3 services.
- Cartographie complète des migrations et des transactions (DB_MIGRATIONS_*, DB_TRANSACTION_PLAN, AUDIT_DB_TRANSACTIONS, TRANSACTION_TESTS_PHASE3).
- Scripts de reset et de cleanup pour la lab DB et la V1.
Ce commit fige l’ensemble du travail de stabilisation P0 (UUID, backend, chat et stream) avant les phases suivantes (Coherence Guardian, WS hardening, etc.).
