**Hi Frappe community,
I’m a backend developer (Python/PostgreSQL) wanting to develop Frappe apps with separate backend/frontend servers like normal Python web development:
- Backend: Python server (8000) with auto-reload
- Frontend: Vite dev server (8080) proxying API calls
- Edit Python → backend restarts
- Edit JS/Vue → frontend hot reloads
**Current setup:**
- Cloned Frappe source: `D:\Dev\frappe\frappe`
- Want Linux/WSL native development
- Goal: Edit→Save→See changes (no bench restart delays)
**Questions:**
- Can I run `bench start` (backend:8000) + `yarn dev` (frontend:8080) separately?
- What’s the `vite.config.js` proxy setup for `/api` → `localhost:8000`?
- Does backend auto-reload Python changes without full `bench restart`?
- How to serve custom app frontend from `app/public` or separate dir?
**Example workflow I’m seeking:****
Terminal 1: bench start # Python API @ localhost:8000
Terminal 2: cd myapp/frontend && yarn dev # Vite @ localhost:8080
Browser: localhost:8080 # Proxies /api → backend
Want this instead of monolithic bench start that bundles everything. Is this the standard modern Frappe dev pattern?
Context: Experienced with Flask/Django + Vite/React separate servers. Frappe’s monorepo feels different.
Thanks!