db_Forge

I get why this feels cleaner, but I’d be nervous about losing the “how did we get here?” part for anything beyond simple schema changes. Current-state SQL is nice for fresh environments, but once real data, backfills, destructive changes, and one-off fixes enter the chat, migration history usually stops being bureaucracy and starts being evidence. A hybrid approach might be the sweet spot: periodically squash into a new baseline, then keep forward migrations from there.