How akasha1618/prompt-craft-studio works
Project type and purpose
• This is a full-stack, serverless Next.js 14 application ("PromptCraft Studio") for end-to-end LLM prompt engineering (creation, chaining, improvement/optimization, testing, analytics) built on TypeScript, Tailwind CSS, Supabase (Auth + PostgreSQL + RLS) and integrating external AI providers (OpenAI, Anthropic).
File structure
• Top-level “app/” folder uses Next.js App Router:
• app/api/… → serverless API routes for prompt operations (generate, improve, optimize, test, user-settings)
• app/auth/… → auth flows (signup, login, reset, callback)
• app/dashboard/… → UI pages for chains, prompts, tests, settings
• app/layout.tsx, app/globals.css → global layout + styling
• Config files: next.config.js, tsconfig.json, tailwind.config.js
• types/database.ts defines Supabase-related types
• README describes Supabase schema (supabase/schema.sql) and deployment guides
README for architectural insights
• Backend: Supabase for Auth & Postgres (with RLS), Next.js API routes as the server layer for AI calls
• AI Integration: external HTTP calls to OpenAI & Anthropic APIs
• Deployment: Vercel (serverless) or manual Node server
Detail without overthinking
• Every major route and flow (prompt creation, chaining, testing, settings)
• RLS security guard at the DB level
• How a new prompt flows from Browser → Next.js API → Supabase → AI → DB → Browser
• Multi-model branching when testing or chaining