Shelfmark + Debrid App Blueprint

This repository now contains a concrete build plan for an app inspired by:

• https://github.com/calibrain/shelfmark
• https://github.com/kikootwo/ReadMeABook

Product goal

Build a self-hostable reading + listening workflow app that can:

1. Track books, reading status, and progress.
2. Ingest from user libraries and watchlists.
3. Resolve downloadable/cached sources via Debrid providers (TorBox first, then others).
4. Send files/streams to reading clients (web reader, Audiobookshelf, Kavita, etc.).

What is included

• docs/product-spec.md: MVP and feature scope.
• docs/system-architecture.md: Backend/services architecture.
• docs/debrid-provider-contract.md: provider adapter interface and normalization rules.
• docs/api-first-endpoints.md: API endpoint draft for building the backend first.

Suggested implementation stack

• Frontend: Next.js + TypeScript + Tailwind.
• Backend API: Fastify (or NestJS) + TypeScript.
• Jobs/Workers: BullMQ + Redis.
• Database: PostgreSQL + Prisma.
• Auth: NextAuth/Auth.js (OAuth + local).
• Storage: S3-compatible + local filesystem fallback.

Next steps

1. Scaffold monorepo (apps/web, apps/api, packages/core).
2. Implement Debrid adapter abstraction and TorBox adapter.
3. Add metadata ingestion (OpenLibrary/Google Books).
4. Add queue-based download + transcoding pipeline.
5. Ship MVP dashboard and library views.