Company Automations Hub

Goal: Host multiple habits for your whole team under a single domain, with per-service subdomains, HTTPS, a visual admin panel, and no manual infrastructure work.

This recipe uses Habits Admin, an all-in-one Docker image that bundles a reverse proxy with auto-HTTPS, DNS management, process supervision, and a web-based admin UI. One docker compose up command gets you from zero to a running multi-tenant habits platform.

What you'll end up with

yourdomain.com              → Admin UI (manage everything from here)
marketing.yourdomain.com    → Cortex instance running your marketing-campaign habit
crm.yourdomain.com          → Cortex instance running your CRM workflow habit
onboarding.yourdomain.com   → Cortex instance running your customer onboarding habit

Each subdomain is fully isolated, HTTPS-secured, and can be started, stopped, or updated independently from the admin panel.

Tools used

Tool	Role
Admin	Orchestrates all Cortex instances, manages DNS, issues TLS certs
Cortex Server	Runs each individual habit in isolation
Base	Optional, mount as a managed service for building new habits
Mirror	Optional, mount as a managed service for habit transfers

Step 1, Pull and start the Docker image

Admin ships as a single Docker image. One compose file starts everything.

docker compose up

What happens when you run this:

• Reverse proxy starts and reads your domain from ADMIN_DOMAIN env. It begins issuing Let's Encrypt certificates automatically for wildcard subdomains.
• DNS server starts and accepts dynamic DNS record creation via the Admin API. No manual zone editing required.
• Process supervisor manages all internal services. If any crashes, it restarts automatically.
• Admin UI starts at https://yourdomain.com and is immediately usable.

You'll need a wildcard DNS A record pointing *.yourdomain.com to your server's IP before starting. The Admin panel handles all subdomain records from there.

Step 2, Assign a subdomain to a habit

Open the Admin UI. In the Services tab, click New Service.

Enter a subdomain name (e.g. marketing) and choose which habit to deploy. The admin panel:

1. Creates a DNS A record pointing marketing.yourdomain.com → server IP
2. Requests a TLS certificate for that subdomain
3. Starts a new Cortex instance on an internal port
4. Routes marketing.yourdomain.com → that Cortex instance

This takes about 5–10 seconds. No SSH, no manual cert commands.

Step 3, Upload or browse habits

In the Library tab you can install habits two ways:

From the built-in showcase library: Browse habits that ship with the Admin image. Click Install and pick which service to deploy it to.

Upload your own .habit file: Drag any .habit binary file into the upload zone. The Admin validates the file, registers it in the library, and makes it available for deployment.

Once a habit is installed to a service, Admin starts the Cortex instance automatically.

Step 4, Your habit is live

Navigate to the subdomain you created. The habit is running, fully isolated, and HTTPS-secured.

Repeat Steps 2–4 for every habit your team needs. Each gets its own subdomain and isolated process. You manage them all from the same admin panel.

Day-2 operations

Updating a habit

Upload a new .habit version in the Library tab. In the Services tab, click Update on the service. Admin performs a zero-downtime swap: starts the new version, waits for it to pass a health check, then stops the old one.

Scaling

Each Cortex instance runs on a single process. For high-traffic habits, you can run multiple Admin instances behind a load balancer, or horizontally scale individual subdomains by running multiple Cortex processes and adding round-robin DNS records.

User management

The Users tab lets you create accounts and assign roles. You can restrict who can access the Admin UI vs. who can only view the services list.

Monitoring

The admin panel shows real-time CPU, memory, and request metrics per service. For deeper observability, Cortex exposes Prometheus-compatible /metrics endpoints that you can scrape with Grafana.

Backup

Admin stores its state (DNS records, service configs, user accounts) in a local SQLite database at /data/admin.db. Mount /data as a Docker volume and snapshot it.

volumes:
  - ./data:/data

Environment variables

Variable	Description	Example
ADMIN_DOMAIN	Root domain for the admin UI and wildcard subdomains	yourdomain.com
ADMIN_PASSWORD	Initial admin UI password	changeme
ACME_EMAIL	Email for Let's Encrypt notifications	admin@yourdomain.com
DNS_API_KEY	Internal DNS API key	secret

Tips for getting the most out of Admin

Use the library as a habit registry. Upload every habit your team uses to the library, even if it's not deployed yet. The library acts as a central registry, you always know which version of a habit is available.

Assign subdomains that mirror your team names. marketing.yourdomain.com, sales.yourdomain.com, ops.yourdomain.com. This makes it obvious to end users which habit they're using without needing a docs page.

Enable Base as a managed service. If anyone on your team needs to edit or build habits, enable the Base managed service in Admin settings. It mounts the visual editor at base.yourdomain.com and connects directly to the library so builders can test and deploy without leaving the browser.

Set up Mirror for habit distribution. Enable the Mirror service in Admin settings. Your team can share new .habit files from Base directly to Admin over a peer-to-peer connection, without email or file sharing services. See the Mirror docs.

Related

• Admin tool reference
• Cortex Server reference
• Mirror tool reference
• No-Code Automation Builder recipe