13 Commits

Author SHA1 Message Date
Hartmut fe79810a85 security: MFA backup codes — issue on enable, redeem at login, regenerate on demand (#43)
CI / Architecture Guardrails (push) Successful in 6m1s
CI / Assistant Split Regression (push) Successful in 6m52s
CI / Lint (push) Successful in 8m40s
CI / Typecheck (push) Successful in 9m45s
CI / Unit Tests (push) Successful in 7m28s
CI / Build (push) Failing after 10m16s
CI / E2E Tests (push) Has been cancelled
CI / Fresh-Linux Docker Deploy (push) Has been cancelled
CI / Release Images (push) Has been cancelled
Adds a one-time-use backup code set so users with a lost authenticator are not
locked out. Codes are Crockford base32 (XXXXX-XXXXX), hashed with argon2id, and
redeemed under a WHERE-guarded delete so a concurrent replay race fails closed.

- New MfaBackupCode model + migration
- Issue 10 codes inside the enable transaction; show plaintext exactly once
- Sign-in page accepts TOTP or backup code, reporting remaining count
- regenerateBackupCodes tRPC mutation wipes + reissues atomically
- Unit coverage for generator, normalizer, verify, redeem, and race path

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-04-17 18:47:18 +02:00
Hartmut 3222bec8a5 security: atomic compare-and-swap for TOTP replay window (#43, part 1)
The previous SELECT → compare → UPDATE sequence let two concurrent login
requests with the same valid 6-digit code both observe a stale lastTotpAt,
both pass the in-JS replay check, and both succeed. A stolen TOTP (shoulder-
surf, phishing-proxy replay) was usable twice within its 30 s window.

Replace the three callsites (login authorize, self-service enable, self-
service verify) with a shared consumeTotpWindow() helper: a single
updateMany() expresses "window unused" as a SQL WHERE clause, so Postgres'
row lock serialises concurrent writers and whichever commits second sees
count=0 and is treated as a replay.

Backup codes (ticket part 2) are tracked as follow-up work.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-04-17 09:11:50 +02:00
Hartmut 93a7fbaa4c security: fail-fast dev-bypass flag in production (#42)
Both auth.ts and trpc.ts now delegate the E2E_TEST_MODE-in-production
check to a single shared helper (packages/api/src/lib/runtime-security.ts).
trpc.ts used to only console.warn; it now throws at module load time,
matching the behaviour already enforced by assertSecureRuntimeEnv on the
auth side. A future refactor can no longer silently drop the guard on
either side.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-04-17 08:56:27 +02:00
Hartmut 64ca79f3a6 ci: add @vitest/coverage-v8 to workspace packages; set REDIS_URL on build
CI / Architecture Guardrails (push) Failing after 14s
CI / Unit Tests (push) Failing after 4m33s
CI / Assistant Split Regression (push) Successful in 7m17s
CI / Build (push) Has been cancelled
CI / E2E Tests (push) Has been cancelled
CI / Typecheck (push) Has started running
CI / Fresh-Linux Docker Deploy (push) Has been cancelled
CI / Release Images (push) Has been cancelled
CI / Lint (push) Has started running
CI unit-test runs vitest run --coverage in each workspace package, but only
apps/web declared the coverage-v8 dep. In pnpm workspaces deps aren't
hoisted across packages, so engine/staffing/api/application/shared need it
directly.

The build job also needs REDIS_URL because collecting page data for
/api/perf imports a module that throws if REDIS_URL is missing under
NODE_ENV=production. A placeholder value satisfies the check (no actual
Redis connection is made at build time).
2026-04-12 18:38:21 +02:00
Hartmut 78d50b78d3 fix: script portability and npm security updates
Scripts:
- stop.sh: replace Linux-only fuser with cross-platform lsof fallback
- start.sh: parameterize port (APP_PORT) and container name (dynamic lookup)
- app-dev-start.sh: cross-platform stat (GNU -c / BSD -f) and setpriv/su fallback
- deploy-compose.sh: parameterize Docker registry via DOCKER_REGISTRY env var
- harden-postgres.sh: make DB_USER and DB_NAME configurable via env vars

NPM security:
- next: 15.5.12 → 15.5.15 (fixes HTTP request smuggling CVE)
- nodemailer: 8.0.1 → 8.0.5 (fixes SMTP command injection CVEs)
- lodash-es: add pnpm override to force >=4.18.0 (fixes code injection + prototype pollution)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-10 14:06:58 +02:00
Hartmut f2d65d3cd4 test(api): add assistant split regression runner 2026-04-01 00:51:23 +02:00
Hartmut 82466a4e34 fix(api): derive secure sse subscriptions 2026-03-30 14:20:18 +02:00
Hartmut 4f48afe7b4 feat(planning): ship holiday-aware planning and assistant upgrades 2026-03-28 22:49:28 +01:00
Hartmut 9d43e4b113 feat: ACN Application Security Standard V7.30 compliance (19/23 items)
CRITICAL — Authentication & Access:
- TOTP MFA: otpauth-based, QR setup UI, sign-in flow integration,
  admin disable override, /account/security self-service page
- Session Timeouts: 8h absolute (maxAge), 30min idle (updateAge)
- Failed Auth Logging: Pino warn for invalid password/user/totp,
  info for successful login, audit entries for all auth events
- Concurrent Session Limit: ActiveSession model, oldest-kick strategy,
  max 3 per user (configurable in SystemSettings)

CRITICAL — HTTP Security:
- HSTS: max-age=31536000; includeSubDomains
- CSP: script/style/img/font/connect-src with Gemini/OpenAI whitelist
- X-XSS-Protection: 0 (CSP replaces legacy)
- Auth page cache: no-store, no-cache, must-revalidate
- Rate Limiting: 100/15min general API, 5/15min auth (Map-based)

Data Protection:
- XSS Sanitization: DOMPurify on comment bodies
- autocomplete="new-password" on all password/secret fields
- SameSite=Strict on all cookies (Credentials-only, no OAuth)
- File Upload Magic Bytes validation (PNG/JPEG/WebP/GIF/BMP/TIFF)

Logging & Monitoring:
- Login/Logout audit entries (Auth entityType)
- External API call logging with timing (OpenAI, Gemini)
- Input validation failure logging at warn level
- Concurrent session tracking in ActiveSession table

Documentation:
- docs/security-architecture.md (11 sections)
- docs/sdlc.md (CI pipeline, security gates, incident response)
- .gitea/PULL_REQUEST_TEMPLATE.md (security checklist)

Schema: User.totpSecret/totpEnabled, SystemSettings.sessionMaxAge/
sessionIdleTimeout/maxConcurrentSessions, ActiveSession model

Tests: 310 engine + 37 staffing pass. TypeScript clean.

Co-Authored-By: claude-flow <ruv@ruv.net>
2026-03-27 14:16:39 +01:00
Hartmut cd78f72f33 chore: full technical rename planarchy → capakraken
Complete rename of all technical identifiers across the codebase:

Package names (11 packages):
- @planarchy/* → @capakraken/* in all package.json, tsconfig, imports

Import statements: 277 files, 548 occurrences replaced

Database & Docker:
- PostgreSQL user/db: planarchy → capakraken
- Docker volumes: planarchy_pgdata → capakraken_pgdata
- Connection strings updated in docker-compose, .env, CI

CI/CD:
- GitHub Actions workflow: all filter commands updated
- Test database credentials updated

Infrastructure:
- Redis channel: planarchy:sse → capakraken:sse
- Logger service name: planarchy-api → capakraken-api
- Anonymization seed updated
- Start/stop/restart scripts updated

Test data:
- Seed emails: @planarchy.dev → @capakraken.dev
- E2E test credentials: all 11 spec files updated
- Email defaults: @planarchy.app → @capakraken.app
- localStorage keys: planarchy_* → capakraken_*

Documentation: 30+ .md files updated

Verification:
- pnpm install: workspace resolution works
- TypeScript: only pre-existing TS2589 (no new errors)
- Engine: 310/310 tests pass
- Staffing: 37/37 tests pass

Co-Authored-By: claude-flow <ruv@ruv.net>
2026-03-27 13:18:09 +01:00
Hartmut fbeab5cd79 feat: Sprint 5 — AI insights, webhooks/Slack, PWA, performance monitoring
AI-Powered Insights (G9):
- Rule-based anomaly detection: budget burn rate, staffing gaps, utilization,
  timeline overruns across all active projects
- AI narrative generation via existing Azure OpenAI integration
- Cached in project dynamicFields to avoid regeneration
- New /analytics/insights page with anomaly feed + project summaries
- Sidebar nav: "AI Insights" under Analytics

Webhook System + Slack (G10):
- Webhook model in Prisma (url, secret, events, isActive)
- HMAC-SHA256 signed payloads with 5s timeout fire-and-forget dispatch
- Slack-aware: routes hooks.slack.com URLs through Slack formatter
- 6 events integrated: allocation.created/updated/deleted, project.created/
  status_changed, vacation.approved
- Admin UI: /admin/webhooks with CRUD, test button, event checkboxes
- webhook router: list, getById, create, update, delete, test

PWA Support (G11):
- manifest.json with standalone display, brand-colored icons (192+512px)
- Service worker: cache-first for static, network-first for API, offline fallback
- ServiceWorkerRegistration component with 60-min update checks
- InstallPrompt banner with 30-day dismissal memory
- Apple Web App meta tags + viewport theme color

Performance Monitoring (A15):
- Pino structured logging (JSON prod, pretty dev) via LOG_LEVEL env
- tRPC logging middleware on all protectedProcedure calls
- Request ID (UUID) per call for log correlation
- Slow query warnings (>500ms) at warn level
- GET /api/perf endpoint: memory, uptime, SSE connections, node version

Fix: renamed scenario.apply to scenario.applyScenario (tRPC reserved word)

Co-Authored-By: claude-flow <ruv@ruv.net>
2026-03-20 06:57:20 +01:00
Hartmut d0f04f13f8 feat: enterprise notification & task management system
Phase N.1 — Data Model:
- Extend Notification model with category, priority, task fields (status, action,
  assignee, dueDate, completedAt/By), reminder fields (remindAt, recurrence,
  nextRemindAt), and targeting metadata (sourceId, senderId, channel)
- Add NotificationCategory, NotificationPriority, TaskStatus enums
- Add NotificationBroadcast model for group notifications
- Shared types with parseTaskAction()/buildTaskAction() helpers

Phase N.2 — API:
- Extend notification router: listTasks, taskCounts, updateTaskStatus,
  createReminder/update/delete/list, createBroadcast/listBroadcasts,
  createTask, assignTask, delete
- Broadcast targeting: resolve recipients by user/role/project/orgUnit/all
- Task-action registry: approve_vacation, reject_vacation, confirm_assignment
- Reminder scheduler: 60s poll interval, recurring support, catch-up on start
- SSE events: TASK_ASSIGNED, TASK_COMPLETED, TASK_STATUS_CHANGED,
  REMINDER_DUE, BROADCAST_SENT

Phase N.3 — AI Assistant:
- 7 new tools: list_tasks, get_task_detail, update_task_status,
  execute_task_action, create_reminder, create_task_for_user, send_broadcast
- execute_task_action dispatches to task-action registry with per-action
  permission checks, marks tasks as completed by AI

Phase N.4 — Frontend:
- Enhanced NotificationBell with task badge, tabs (All/Tasks/Reminders)
- TaskCard component with priority badges, due dates, action buttons
- ReminderModal for creating/editing personal reminders
- BroadcastModal for targeted group notifications (manager+)
- NotificationCenter full-page with 5 tabs and bulk actions
- TaskWidget dashboard widget showing open tasks
- Admin broadcast management page
- AppShell nav links for Notifications and Broadcasts
- SSE hook handlers for task/reminder events

Phase N.5 — Auto-Tasks:
- Vacation create → APPROVAL tasks for all managers
- Vacation approve/reject → mark approval tasks as DONE
- Demand create → TASK for managers to fill staffing needs

Co-Authored-By: claude-flow <ruv@ruv.net>
2026-03-18 11:51:49 +01:00
Hartmut dd55d0e78b chore(repo): initialize planarchy workspace 2026-03-14 14:31:09 +01:00