Deploy with an agent
Copy agent prompt
Create the database
- Connect your AWS account and select a project. Create a private SQL database in the same project as the web server.
- The tested setup used
us-east-1, single-AZ Amazon RDS PostgreSQL18.3,db.t4g.micro, 20 GiB of gp3 storage and seven days of automated backups. Use the regional catalog to check availability. Create an empty database namedwikiwith a dedicated database user. - Wait for the database workflow to succeed. Open Connect and use the private writer endpoint and its credentials.
Deploy the web server
Create a web server with these settings.
Set these runtime variables before you deploy.
The tested deployment used the Amazon RDS us-east-1 Root CA RSA2048 G1. Check your database’s CA identifier in the RDS console and select the matching root from the bundle for your Region. For
DB_SSL_CA, remove the PEM header, footer and line breaks from that single certificate. AWS RDS certificate bundles · Wiki.js TLS implementation
Complete setup and verify
- Wait for the web deployment to succeed in History. Open the managed HTTPS URL, create the first administrator and set the site URL to that HTTPS address. Wiki.js handles its schema initialization.
- Before setup completes,
/healthzcan return the setup page with HTTP 200. After setup, check that it returns HTTP 200 with{"ok":true}. - Create and edit a test page. Upload a sample asset. Replace the web task and confirm that the page and asset still load.