- Framework: Next.js 16 (App Router, Turbopack, React 19)
- Language: TypeScript 5.8
- Architecture: Edge Proxy (Stateless, no DB). Session tokens encrypted with AES-256-GCM.
- Styling: Tailwind CSS v4 + Vanilla CSS tokens (
globals.css) - Icons: Native inline SVG components (
src/components/ui/icons.tsx— 57 primitives), zero external icon dependencies. - Testing: Native Node.js Test Runner (
node:test) + Playwright E2E + Agent-as-Judge AI suite + Challenger stress harnesses.
| # | Feature | Description | Milestone | Source |
|---|---|---|---|---|
| 1 | Automated Test Suite (331 tests) | Execute and verify all 331 unit/integration tests across 54 test suites with 0 failures | M1 | ORIGINAL_REQUEST R2 |
| 2 | TypeScript Compilation (0 errors) | Full AST typecheck across all src, test, and script files with npx tsc --noEmit |
M1 | ORIGINAL_REQUEST R2 / Criteria |
| 3 | ESLint Conformance (0 errors/warnings) | Zero ESLint warnings or errors across the repository with npm run lint |
M1 | ORIGINAL_REQUEST R2 / Criteria |
| 4 | Agent-as-Judge AI Evaluation | 9/9 AI tool execution and parameter validation tests passing via scripts/agent-as-judge.ts |
M1 | ORIGINAL_REQUEST R2 / Criteria |
| 5 | Production Turbopack Build | Clean Next.js 16 Turbopack production compilation via npm run build |
M1 | ORIGINAL_REQUEST R2 / Criteria |
| 6 | 11 Dashboard Routes E2E Verification | Clean navigation, zero console errors, zero React exceptions across all 11 modules in Demo Mode | M2 | ORIGINAL_REQUEST R1 |
| 7 | UI & Accessibility Compliance | WCAG 2.2 AAA contrast (≥7.1:1), 44px min touch targets, rounded corner containment, photo fallbacks | M2 | ORIGINAL_REQUEST R1 |
| 8 | Playwright E2E Test Suite | 13/13 E2E tests passing in Playwright test runner (npx playwright test) |
M2 | ORIGINAL_REQUEST R2 |
| 9 | Forensic Integrity & Multi-Perspective Audit | Independent verification by Reviewers, Challengers, and Forensic Auditor | M3 | Audit & Acceptance Criteria |
| # | Name | Scope | Dependencies | Status |
|---|---|---|---|---|
| 1 | Automated Verification Suites Baseline | TypeScript (0 errors), Lint (0 errors), Unit Tests (331/331), Agent-as-Judge (9/9), Build | none | DONE |
| 2 | End-to-End Browser & UI Module Audit | 11 Dashboard Routes in Demo Mode, zero console errors, zero hydration errors, Playwright E2E | M1 | DONE |
| 3 | Multi-Perspective Review & Forensic Audit | Dual Reviewers, Dual Challengers, and Forensic Auditor verification | M1, M2 | DONE |
- All 11 routes (
attendance,timetable,marks,profile,fee,exam-seating,circulars,hostels,library,tools,copilot) load cleanly with HTTP 200–399 in demo mode. - 0 unhandled console errors or React hydration warnings emitted.
- All interactive controls respect
min-h-[44px] min-w-[44px]for WCAG 2.2 AAA.
npx tsc --noEmitexits with 0 errors.npm run lintexits with 0 errors.npm testexecutes 331 tests across 54 suites with 100% pass rate.npx tsx scripts/agent-as-judge.tsexecutes 9 AI evaluations with 100% pass rate.npm run buildgenerates 17 static pages and 8 dynamic endpoints cleanly.npx playwright testpasses 13/13 E2E test specs.
src/app/— Next.js App Router pages and API routessrc/components/ui/icons.tsx— Native zero-dependency SVG icon system (57 primitives)src/components/— UI components and widgetssrc/lib/scrapers/— ERP web scrapers and data parserssrc/lib/fixtures/— Demo mode mock data fixturessrc/lib/session.ts— AES-256-GCM session encryptiontests/&src/**/*.test.ts— Native Node.js test runner unit & integration suitesscripts/— Challenger stress harnesses, agent-as-judge evaluation, browser audite2e/— Playwright end-to-end test specifications