rla-web-cms

Payload CMS 3 / Next.js app that is becoming the single, unified content-management platform for Rehwaldt Landschaftsarchitekten (RLA), consolidating data that today lives scattered across several disjointed legacy systems (a WordPress intranet, a PHP public website, and a separate time-tracking API). It is early-stage: most collections exist and an ETL pipeline pulls in real legacy data, but this is not yet the system of record for anything — it's being built up collection by collection.

See CLAUDE.md for the full architecture, consolidation plan, and the legacy systems this project is absorbing data from.

Stack

• Payload CMS 3 on Next.js 16
• Postgres via @payloadcms/db-postgres (not Mongo)
• Localization: de (default) / en, with fallback
• Leaflet for the custom map/location field

Local development

1. Copy the example environment file and fill in real values:

   cp .env.example .env

   • DATABASE_URL — a Postgres connection string
   • PAYLOAD_SECRET — any random string
   • WP_BASE_URL, LEGACY_DB_* — only needed to run the ETL scripts (see below), not for general development

2. Install dependencies and start the dev server:

   pnpm install
   pnpm dev

3. Open http://localhost:3000/admin and follow the on-screen instructions to create your first admin user.

Docker (optional)

docker-compose.yml spins up a local Postgres instance alongside the app for local development (not the same thing as the standalone Dockerfile, which is a separate production-image build not wired into this compose file):

docker compose up

Point DATABASE_URL in .env at postgresql://payload:payload@postgres:5432/payload (or whatever POSTGRES_USER/POSTGRES_PASSWORD/POSTGRES_DB you set in docker-compose.yml) to use it.

Scripts

Command	Purpose
pnpm dev / pnpm build / pnpm start	Standard Next.js dev/build/start
pnpm generate:types	Regenerates src/payload-types.ts from the collection configs — run after any collection/field change
pnpm generate:importmap	Regenerates Payload's admin import map
pnpm lint	ESLint
pnpm test	Runs both test:int (Vitest) and test:e2e (Playwright)

ETL pipeline

Legacy data (WordPress, MySQL) is migrated into Payload via scripts in src/scripts/, one folder per source:

Command	Source
pnpm etl:employees	WordPress mitarbeiter CPT + legacy WerkX MySQL HR data → Employee
pnpm etl:news	WordPress news CPT → News
pnpm etl:werkx:projects / pnpm etl:werkx:contracts	rla-werkx-api's own exported project/contract data → Projects / Employee WerkX contracts

See CLAUDE.md's "ETL pipeline" section for the full extract/transform/load shape these follow, and issue tracking in this repo's GitHub issues for which collections still lack an ETL entirely.

Collections

Users, Media, News, Awards, Publications (+ Publication Categories), Jobs, Employee, Office Location, Mines (+ Mine Features), Themes, and Projects (+ Project Phases / Partials / Services / Flags / Categories) — see CLAUDE.md for what each one models and the legacy system it's replacing.