Configuration
The default file is pyahu.yaml. The CLI searches for this file from the current directory upward. Use --file or -f when you want to point to a different path.
Minimal example
Section titled “Minimal example”apiVersion: cli.pyahu.io/v1alpha1kind: Stackmetadata: name: local-devservices: postgres: enabled: true ports: primary: 5432 databases: - name: appComplete stack
Section titled “Complete stack”apiVersion: cli.pyahu.io/v1alpha1kind: Stackmetadata: name: platformservices: postgres: enabled: true zitadel: enabled: true externalURL: https://zitadel.localhost rabbitmq: enabled: true kafka: enabled: true kafkaConnect: enabled: true connectors: - name: app-cdc type: source kind: debezium.postgres tables: include: - public.orders kafkaUI: enabled: trueTCP services expose a host port within the service that owns the endpoint:
services: postgres: ports: primary: 15432 kafka: ports: bootstrap: 19092The HTTP UIs (ZITADEL, RabbitMQ, Kafka UI) do not use a host port: they go
through Traefik on 80/443 with *.localhost hostnames (https://kafka-ui.localhost,
https://rabbitmq.localhost, https://zitadel.localhost). To change the ZITADEL
domain, use services.zitadel.externalURL.
Do not use cluster.ports in presets or new documentation. The CLI keeps silent compatibility with this legacy format, but it is not the v1 surface.
Global config
Section titled “Global config”You can define global defaults in the system configuration directory:
| System | Typical path |
|---|---|
| Linux | ~/.config/pyahu.yaml |
| macOS | ~/Library/Application Support/pyahu.yaml |
The global file is loaded first; the project’s pyahu.yaml overrides the values.
Local credentials
Section titled “Local credentials”PostgreSQL, Zitadel, and RabbitMQ credentials can live in the local pyahu.yaml or in the global config.
For PostgreSQL, changing the password after the volume already exists updates the Secrets and the CLI output, but the user inside the database may keep the old password. For local rotation, recreate the cluster or alter the role inside PostgreSQL.