Employees (internal admin/staff) had no self-service password reset —
only /login, /logout, /session existed. Adds /api/admin/auth/forgot-password
and /api/admin/auth/reset-password, mirroring the existing users-table flow
but against EmployeeRepository and a distinct Redis key namespace
(reset:employee:*) so a code for one identity store can never be consumed
against the other.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The admin/employee login handler had no brute-force protection, unlike
the regular user login path. Given these accounts hold internal/
super-admin privileges, add a tighter limit (5 attempts/15min vs 10
for regular users) using the existing sliding-window Redis limiter.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
- Update jsonwebtoken from 9.3 to 10.3 in crates/auth/Cargo.toml and crates/contracts/Cargo.toml
- Create .cargo/audit.toml to ignore false positives for local workspace crates 'cache' and 'users'
- Fix pre-existing compile errors in crates/cache/src/ollama.rs (missing reqwest dep, broken format! string literals)
- Add reqwest workspace dependency to crates/cache/Cargo.toml
- Add AI credit management endpoints for companies
- Add AI usage history tracking
- Add AI content generation with Ollama integration
- Add Ollama client for generating job descriptions, resume analysis, and cover letters
- Integrate AI router into companies service
- Generate 6-digit code instead of UUID token for password reset
- Store in Redis with 15 min TTL (was 1 hour)
- Update email template to show code instead of reset link
- Update ResetPasswordPayload to accept code instead of token
- Update send_password_reset_email to accept code parameter
- Add cache::ai module with Redis rate limiting for AI generations
- Add functions: check_ai_rate_limit, get_ai_usage, cache_ai_response,
get_cached_ai_response, invalidate_ai_cache, reset_daily_usage
- Update check_and_increment_usage to use Redis fast-path before DB
- Redis key pattern: ai:rate:{user_id} for 24hr sliding window counter
- Add crates/cache with client, otp, rate_limit, token, lead, jobs modules
- OTP tokens stored in Redis (15-min TTL, single-use GETDEL on verify)
- Refresh tokens stored in Redis (30-day TTL) — removed DB storage
- Password reset tokens stored in Redis (1-hour TTL, single-use)
- Rate limiting: register (10/hr), login (10/15min), OTP resend (3/hr), lead (5/hr), job post (20/hr)
- Lead request deduplication: 24-hour Redis lock per professional+requirement pair
- Marketplace listings cached in Redis (5-min TTL per profession+page+limit)
- Add ProfessionState{pool, redis} to contracts crate, replacing bare PgPool in all 9 profession apps
- All profession handlers and main.rs updated to use ProfessionState
- REDIS_URL env var (default: redis://127.0.0.1:6379) used across all services
- Fix profession model struct name mangling in 6 handlers (MakeupArtistRepository etc.)
- Add custom_data JSONB migration for all 9 profession profile tables
- Add onboarding_state model and repository (save_progress, complete, is_complete)
- Add onboarding handler accepting roleKey:String (not role_id:UUID) for frontend compat
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
