Overview

A recipe formatting and publishing tool that takes structured recipe data and outputs it as Markdown, a printable PDF, or an SEO-optimized YouTube video description.

🚀 Live Demo

View Live Deployment (Placeholder) | Developer Portfolio

---

🛑 Project Status: Scoping & Architecture Phase

This repository is currently in the Product Specification & Architecture Design phase. The feature scope, technical stack choices, and data boundaries have been mapped below. Implementation is scheduled to begin next week.

Last Updated: August 29, 2026

---

🗺️ Product Roadmap & Milestones

This project follows an agile, phased release strategy. By decoupling core functionality from complex integrations, the roadmap minimizes risk, accelerates time-to-market for the MVP, and ensures engineering stability at every milestone.

┌─────────────────────────────┐ ┌─────────────────────────────┐ ┌─────────────────────────────┐
│     Phase 1: Foundation ──> │ │    Phase 2: AI Engine   ──> │ │    Phase 3: Ecosystem       │
│    CRUD & Local Exporters   │ │    OpenAI & Formatting      │ │   Analytics & Social Sync.  │
└─────────────────────────────┘ └─────────────────────────────┘ └─────────────────────────────┘

🟩 Phase 1: Foundation & Local Compilation (MVP) — Current Phase

Objective: Establish core data contracts, state management, and basic local export utilities without external API dependencies.

• Data Architecture: Design and freeze the strict JSON schema for single-serving recipes (macros, ingredients, custom conversion units).
• Core UI Layer: Build the interactive recipe input wizard and state management engine in SvelteKit.
• Local Markdown Exporter: Write the client-side JavaScript utility to map raw JSON strings into clean, downloadable .md files.
• Local PDF Engine: Implement the serverless HTML-to-PDF rendering pipeline for instant, print-ready document streaming.

🟨 Phase 2: AI Orchestration & Smart Copy (V1.1)

Objective: Safely integrate generative AI to handle creative formatting workloads while maintaining strict data constraints.

• LLM Pipeline Integration: Connect the Next/Svelte serverless API routes to the OpenAI SDK using structured JSON outputs.
• YouTube SEO Engine: Develop the system prompts required to translate raw recipe steps into highly engaging, timestamped YouTube descriptions.
• Error & Rate Limit Handling: Implement robust client-side error states and server-side fallback text routines in case of third-party API timeouts or token limits.

🟦 Phase 3: Ecosystem Expansion & Analytics (V2.0)

Objective: Transition the app from a single-user workflow tool into a full-scale portfolio asset with user retention features.

• Etsy & Analytics Dashboard: (For the Crochet App sister project) Add simple tracking to monitor download success rates and export metrics.
• Cloud Persistence Layer: Integrate a lightweight database (e.g., Supabase or PostgreSQL) to allow creators to save, edit, and history-track past generated outputs.
• One-Click Clipboard Optimization: Introduce an interactive "Preview & Tweak" canvas allowing users to modify AI outputs before final copying.

---

Feature Scope

                  ┌──────────────────────┐
                  │   Svelte Frontend    │
                  └──────────┬───────────┘
                             │ (Recipe Input & Formatting Options)
                             ▼
                  ┌──────────────────────┐
                  │  Node.js API Route   │
                  └──────────┬───────────┘
                             │
         ┌───────────────────┼───────────────────┐
         ▼                   ▼                   ▼
┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
│  LLM API (OAI)  │ │  PDFKit /       │ │   Markdown      │
│ Gen Description │ │  Puppeteer      │ │   Formatter     │
└─────────────────┘ └─────────────────┘ └─────────────────┘

• Frontend: Uses Svelte.js with Tailwind CSS
• Backend: Uses Node.js with Express
• Authentication: Standard JWT with username/password
• Data Structure: Store your raw recipe structured as JSON (e.g., fields for title, description, ingredients[], instructions[], and macros)

The Markdown Exporter

This requires no external APIs. Write a simple pure-JavaScript utility function that maps your recipe JSON template directly into a standard strings-based .md file format.

The PDF Generator

Use a lightweight server-side library like PDFKit or Puppeteer to render an HTML preview of the recipe directly into a clean, printable PDF download stream.

The YouTube Optimizer

Integrate a structured LLM API call (like OpenAI or Anthropic). Pass the recipe JSON to the model with a strict system prompt instruction:

"Act as a YouTube SEO expert. Turn this recipe JSON into an engaging video description, complete with chapters/timestamps placeholders, a hook, and relevant hashtags."