2 - Capstone

This capstone proves the four claims from Module 2 in real time: • A focused R-I-C-E prompt produces a usable PRD in minutes — not weeks — with user stories, Gherkin acceptance criteria, and an explicit out-of-scope list. • AI generates a normalized schema with indexes, constraints, and seed data from a plain-English domain description — not a whiteboard session. • Mermaid.js plus AI produces architecture, sequence, and ER diagrams in seconds, in a syntax that lives in your repo and renders inside your app. • All three artifacts plug directly into a running full-stack application — the diagrams are not just docs in a wiki, they are a feature on the page. You will go from blank directory to a running TaskFlow Mini app — Express backend, SQLite database with the AI-designed schema, and a single-page frontend that renders the AI-generated Mermaid diagrams live in the browser — in one session.