It uses a strict Intermediate Representation (IR) to unify models and detect information loss. Right now it supports: - ERDPlus (old/new) ↔ IR - IR → PostgreSQL DDL - IR → Prisma schema - Loss report when features don’t map (e.g. CHECK constraints in Prisma)
GitHub: https://github.com/tobiager/Erdus Live demo: https://erdus-inky.vercel.app/
Would love feedback from the HN community.