# Production Incident Checklist

Use this checklist before treating a production deploy as healthy again.

## Verify routing and HTML

1. verify `/health`
2. verify `/`
3. verify `/index.html`
4. verify nginx can still read the frontend dist path

## Verify auth contract

1. verify `POST /services/oauth2/token` returns `200`
2. verify token `instance_url` is on the public production domain
3. verify token `id` is also on the public production domain
4. verify browser login in a clean context

## Verify runtime authority

1. verify the generated backend runner and env file agree on public URLs
2. verify the systemd service is running the expected runner
3. verify the installed standard pack source still points at the rebuilt local final pack