Ngrok Configuration File Standards

This rule applies to files matching the patterns above.

# Ngrok Configuration File Standards

## Rule
All ngrok setups MUST use a configuration file (ngrok.yml) instead of CLI flags. Configuration MUST be consistent across the team.

## Format
```yaml
version: "3"
agent:
  authtoken: ${NGROK_AUTHTOKEN}

tunnels:
  tunnel-name:
    addr: service:port
    proto: http
    domain: descriptive-name.ngrok.dev
    inspect: true
```

## Naming Conventions
| Item | Convention | Example |
|------|-----------|---------|
| Tunnel name | kebab-case, matches service | `web-app`, `api-server` |
| Domain | project-env.ngrok.dev | `myapp-dev.ngrok.dev` |
| Config file | ngrok.yml or ngrok-{env}.yml | `ngrok-dev.yml` |

## Required Settings
```yaml
tunnels:
  web:
    addr: 3000
    proto: http
    inspect: true       # REQUIRED: enable debugging
    domain: app.ngrok.dev  # RECOMMENDED: stable URL
```

## Multiple Environments
```yaml
# ngrok-dev.yml — for local development
tunnels:
  app:
    addr: 3000
    proto: http
    domain: myapp-dev.ngrok.dev

# ngrok-staging.yml — for staging exposure
tunnels:
  app:
    addr: 3000
    proto: http
    domain: myapp-staging.ngrok.dev
    oauth:
      provider: google
      allow_domains: ["company.com"]
```

## Examples

### Good
- Configuration file in project root (gitignored authtoken)
- Descriptive tunnel names matching services
- Custom domains for stable URLs
- inspect: true for debugging

### Bad
```bash
# CLI flags instead of config file
ngrok http 3000 --domain=my-app.ngrok.dev --inspect

# No one remembers the exact flags to use
```

## Enforcement
Include ngrok.yml template in project setup documentation.
Add authtoken patterns to .gitignore.
Review tunnel configurations in team onboarding.