Full audit (bugs/UI-UX/backend/feature-gap) turned up a money-leak AI quota
bypass, webhook SSRF, and a long tail of missing pagination/auth/a11y work.
Fixes land together since HANDOFF.md tracked them as one backlog.
- AI routes charge tier quota before generating; nutrition POST is author-only
- Webhook dispatch re-validates URL per delivery (SSRF/DNS-rebinding), treats
redirects as failures; recipe.published now actually dispatches
- New indexes/unique constraints on recipes, meal-planning, comments FK cascade
- Recipe PUT/restore snapshot only inside the transaction, after validation
- Recipe DELETE cleans up S3 objects (recipe + review photos)
- Optimistic UI (favorite/star/follow/shopping-list) rolls back on failure
- Upload presign enforces file size cap + per-tier storage quota
- Route-level loading/error/not-found states across (app), admin, and root
- middleware.ts guards (app)/admin; requireAdmin checks DB role, not cached
session; rate limiting applied to both session and API-key branches,
bucketed per key; Stripe webhook dedupes by event id
- Pagination added to recipes, feed, profile, comments, pantry, admin tables
- Nav shows real avatar + profile link + dark-mode toggle; destructive actions
standardized on AlertDialog
- Unsaved-changes guard + real ingredient/step validation on recipe form;
canonical /recipes/[id] used in-app; next/image migration; aria-labels and
alt text across icon buttons, avatars, recipe photos
- packages/api-types removed (zero callers, too drifted to safely rewire);
openapi.ts and ai-keys error shape drift fixed; BYOK decrypt failures now
surface instead of silently falling back to the platform key
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
- user_blocks table (composite PK), Block/Unblock button on profiles.
Blocking severs any existing follow relationship both ways and
prevents the blocked party from following, commenting on the
blocker's recipes, or the blocker from seeing their comments.
- reports table (recipe/comment/user targets, pending/reviewed/
dismissed status). ReportButton on comments, admin review queue at
/admin/reports with dismiss/mark-reviewed actions, audit-logged.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
- New invites table: token-gated signup, optional email lock,
role/tier override, single-use, expiry.
- SIGNUPS_DISABLED site setting toggle at /admin/settings.
- databaseHooks.user.create gate in auth/server.ts blocks new account
creation (email + Google OAuth) when disabled unless a valid invite
cookie is present; applies invite role/tier and marks it consumed.
- /admin/invites: create/list/revoke shareable invite links.
- /admin/users: "Create user" dialog — admin sets email/role/tier,
account is pre-verified, user gets a set-password email (admin
never sees a password).
- Signup page reads ?invite=, validates via public
/api/v1/invites/[token], locks the form when signups are closed
and no valid invite is present.
- proxy.ts: allowlist /api/v1/invites/ for anonymous invite checks.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Use -1 as a sentinel meaning "no cap" in tier_definitions numeric
columns. checkAndIncrementTierLimit skips the usage-cap WHERE clause
when the limit is -1. Admin tier UI gets an Unlimited switch per
field that disables the input and sends -1.
- ai-generate-dialog: useLocale() returns {locale,setLocale} object, not
string — was stringifying whole object as default language value
- auth/server: add trustedOrigins so session cookie survives reverse-proxy
deployments where BETTER_AUTH_URL differs from container's own view
- admin: add tier limit editor (/admin/tiers) and per-user usage reset
button, both audit-logged
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Sticky sidebar with back-to-app link. Overview: user counts, recipe photos, AI calls/month.
User management with role/tier editing. Audit log (all admin actions).
Storage page with photo breakdown by tier. AI config showing DB vs .env key source.
Site settings page to override .env values at runtime (encrypted in DB).
Arnaud