Papermark

# MISSION: Scaffold "Papermark" Production Architecture
You are an expert Senior Staff Software Architect and Full-Stack Engineer. Your mission is to scaffold and implement a production-grade, highly reliable, and modular codebase following the proven architecture of **Papermark**.
---
## 1. PROJECT SPECIFICATIONS & BENCHMARK
- **Reference Architecture**: Papermark
- **What It Does**: Papermark is an open-source, self-hostable document-sharing platform (a DocSend alternative) built on Next.js and PostgreSQL, with secure links, custom domains, and document analytics.
- **Domain & Category**: Document Analytics SaaS
- **Production Scale**: Solo
- **Development Mode**: AI-AGENT
- **Architectural Rationale**: Splitting enterprise-only code into an ee/ directory keeps the open-source core unambiguous, while pluggable S3-or-Vercel-Blob storage, a separate Tinybird analytics pipeline, and Trigger.dev-managed background jobs let the same Next.js app scale from a Vercel deployment to a self-hosted one.
- **Live Website Reference**: https://papermark.io
- **Source Repository**: https://github.com/mfts/papermark
---
## 2. PRODUCTION TECH STACK
- **Full Stack Array**: TypeScript, Next.js, React, Prisma, PostgreSQL, Tailwind CSS, NextAuth.js, Stripe, Resend, Vercel, S3, Zod, Tinybird
- **Primary Language(s)**: TypeScript, JavaScript, CSS
- **License of the reference repo**: NOASSERTION
- **Frontend**: Next.js — Runs both the legacy Pages Router (pages/) and the App Router (app/(auth), app/(ee)) side by side.; React — Component layer for both routers, on React 18.; Tailwind CSS — Utility CSS with tailwindcss-animate and forms plugins.; Radix UI — Headless primitives (@radix-ui/react-dialog, -dropdown-menu, -tabs, and more) underlying the component library.
- **Backend & APIs**: Next.js — API routes under pages/api and app/api implement the backend; vercel.json tunes memory for two of them.; NextAuth.js — Handles authentication, including a passkeys provider and SAML SSO via @boxyhq/saml-jackson.; Trigger.dev — Runs background jobs, including Python-based tasks via the @trigger.dev/python package.; mupdf — Server-side PDF rendering and annotation; two API routes get 2048MB Vercel function memory for it.
- **Data & persistence**: PostgreSQL — Primary relational database, per the README and the postgresql topic on the repo.; Prisma — ORM for the Postgres schema; the schema is split into multiple files under prisma/schema instead of one file.; S3 — One of two supported blob storage backends via @aws-sdk/client-s3 and @aws-sdk/lib-storage.; Vercel Blob — Alternative blob storage backend via @vercel/blob, selectable instead of S3.; Redis — @upstash/redis and @upstash/ratelimit back caching and API rate limiting.
- **Infrastructure & deploy**: Vercel — vercel.json configures per-function memory; vercel-build script runs prisma migrate deploy before next build.; Tinybird — Separate real-time analytics service; the Pipfile pins tinybird-cli to push datasources from lib/tinybird.; Stripe — Billing integration, used in both the core app and the ee/stripe enterprise module.
- **Tooling, testing & ops**: OpenAI API — The AI SDK (ai, @ai-sdk/openai, @ai-sdk/google-vertex) and openai package power AI features such as document Q&A.; Model Context Protocol — @modelcontextprotocol/sdk is a direct dependency, exposing or consuming MCP tools.
---
## 3. KEY ARCHITECTURAL DECISIONS (audited from https://github.com/mfts/papermark @ ed19717)
1. **Pages Router and App Router coexist in the same Next.js app**: The tree has both a pages/ directory (pages/_app.tsx, pages/api, pages/dashboard.tsx) and an app/ directory with route groups app/(auth) and app/(ee), so newer features migrate to the App Router while older routes stay on the Pages Router rather than a big-bang rewrite.
2. **Enterprise features live in an isolated ee/ directory with its own license**: ee/LICENSE.md, ee/features, ee/limits and ee/stripe separate paid-tier code (billing limits, Stripe integration) from the core app, an open-core structure that keeps the base product’s license unambiguous.
3. **Blob storage is pluggable between S3 and Vercel Blob**: package.json depends on both @aws-sdk/client-s3/@aws-sdk/lib-storage and @vercel/blob, and the README lists "AWS S3 or Vercel Blob" as the two supported options, so deployments choose a backend via environment configuration instead of a hardcoded provider.
4. **Document analytics run through Tinybird, not the Postgres database**: A dedicated Pipfile pins tinybird-cli and lib/tinybird holds datasource/endpoint definitions pushed with `tb push`, and the README’s "Tinybird Instructions" section documents this as a separate setup step from the Prisma/Postgres database, keeping high-volume view-event analytics off the primary OLTP store.
5. **Prisma schema is split into multiple files, not one schema.prisma**: package.json sets "prisma": { "schema": "prisma/schema" }, pointing Prisma at a directory (prisma/schema) rather than the conventional single schema.prisma file, which lets large models be organized across several files.
6. **Background and Python jobs both run through Trigger.dev v4**: package.json depends on @trigger.dev/sdk, @trigger.dev/react-hooks, and @trigger.dev/python together with a trigger.config.ts at the repo root, so both TypeScript and Python background tasks are scheduled through one job platform instead of a custom queue.
---
## 4. NON-NEGOTIABLE ARCHITECTURAL GUARDRAILS
1. **Monorepo & Modular Separation**:
   - Structure as a Turborepo monorepo with strict package boundaries:
     - `apps/web`: Application UI, routing, layouts, and server endpoints.
     - `packages/ui`: Shared design tokens, CSS variables, and Radix UI primitive components.
     - `packages/db`: Database schemas, client singleton, declarative migrations, and seed scripts.
     - `packages/config`: Shared TypeScript, ESLint, and build configurations.
2. **Strict Type Safety & Zero `any` Policy**:
   - Enable `strict: true`, `noImplicitAny: true`, and `strictNullChecks: true`.
   - Validate ALL external inputs, API request bodies, and query parameters with **Zod** schemas before execution.
3. **Frontend & Rendering Guidelines**:
   - Utilize Next.js 15 App Router: default to React Server Components (RSC) for data fetching. Handle mutations through Server Actions with Zod validation. Keep 'use client' directives restricted strictly to interactive leaf components.
4. **Design System & Aesthetics**:
   - Keep every color, radius, shadow and font in a single token file (CSS variables) and consume tokens everywhere; never hardcode hex values in components.
   - Prefer crisp 1px borders and one subtle shadow scale over blurry default shadows. Pair one sans-serif for body/headings with one monospace for tags, badges, metrics, and code.
5. **Data Layer & Reliability**:
   - Write declarative schema definitions with foreign keys, composite indexes on queried filters, and automated timestamp triggers.
   - Use connection pooling and prepared statements for serverless database execution.
---
## 5. STEP-BY-STEP SCAFFOLDING ROADMAP
- **Phase 1: Workspace & Root Config**: Initialize package manager, monorepo configuration (`turbo.json`, `tsconfig.base.json`, `package.json`).
- **Phase 2: Database Schema & Client**: Set up the data layer (PostgreSQL, Prisma, S3, Vercel Blob, Redis): client, connection pool, models, and migration scripts.
- **Phase 3: Design Tokens & UI Primitives**: Build accessible `Button`, `Input`, `Card`, `Badge`, and layout wrappers inside `packages/ui`.
- **Phase 4: Core Application Routes & Handlers**: Implement primary authentication, user session handling, and application routes.
- **Phase 5: Quality Assurance & Build Verification**: Run `tsc --noEmit`, ESLint, Prettier, and smoke test suites to ensure zero compilation or runtime errors.
---
## 6. EXECUTION INSTRUCTIONS
1. Review all specifications, architectural guardrails, and stack choices above.
2. Present the full monorepo directory tree structure.
3. Systematically generate the complete, production-ready codebase according to the 5-phase roadmap above — starting with the root workspace setup, followed by the database schema, UI design system package, and full-stack application routes until the repository is fully scaffolded and ready to run.