# Auto-configuration reference

When you enable `spec.frontend.generatedConfig` in your `nais.yaml`, the platform generates configuration values your frontend can use at runtime. This removes the need to hardcode collector URLs or app metadata.

```yaml
spec:
  frontend:
    generatedConfig:
      mountPath: /usr/share/nginx/html/js/nais.js
```

> **`@nais/apm` reads these automatically**
> If you instrument with [`@nais/apm`](/observability/apm/tutorials/track-frontend-errors.md),
> you don't consume this file by hand. `init()` resolves the app name, version,
> environment, and collector URL from Nais meta tags in your served HTML or from
> the build-time `NAIS_*` environment variables — see
> [Configuration resolution](/observability/apm/reference/apm-client-api.md#configuration-resolution).
> This reference matters when you wire up **raw Faro** yourself.

## Generated JavaScript file

The platform creates a JavaScript file at the specified `mountPath` with this structure:

```js
export default {
  telemetryCollectorURL: 'https://telemetry.nav.no/collect',
  app: {
    name: 'my-app',        // from metadata.name in nais.yaml
    namespace: 'my-team',  // from metadata.namespace in nais.yaml
    version: '2024-03-15-abc1234', // extracted from your container image tag
  },
};
```

| Field | Source |
| --- | --- |
| `telemetryCollectorURL` | Set per cluster by the platform operator |
| `app.name` | `metadata.name` from your `nais.yaml` |
| `app.namespace` | `metadata.namespace` from your `nais.yaml` |
| `app.version` | Tag from your container image |

The collector URL is set automatically based on which cluster your app runs in — you don't need separate config for dev and prod.

## Environment variable

The platform also sets this environment variable in your pod:

| Variable | Description |
| --- | --- |
| `NAIS_FRONTEND_TELEMETRY_COLLECTOR_URL` | The telemetry collector URL for your cluster |

This is useful for server-rendered applications (Next.js, Remix) where the server can read the env var and pass it to client-side code. See the [Next.js guide](/observability/frontend/how-to/setup-nextjs.md) for an example.

> **`NEXT_PUBLIC_` env vars are build-time only**
> Don't re-export this as a `NEXT_PUBLIC_*` variable. Next.js inlines `NEXT_PUBLIC_*` values at `next build` time, so they won't reflect the runtime pod environment.

## Collector URLs by environment

| Cluster | Collector URL |
| --- | --- |
| `prod-gcp` | `https://telemetry.nav.no/collect` |
| `dev-gcp` | `https://telemetry.ekstern.dev.nav.no/collect` |

On-premises clusters (`prod-fss`, `dev-fss`) are not supported.

## When to use auto-configuration

**Static frontends (nginx, CDN):** Use the generated `nais.js` file. Import it at runtime and use `nais.telemetryCollectorURL` and `nais.app`. See the [setup guide](/observability/frontend/how-to/setup-faro.md#auto-configuration).

**Server-rendered apps (Next.js, Remix):** Read `NAIS_FRONTEND_TELEMETRY_COLLECTOR_URL` on the server and pass it to your client-side Faro initialization. See the [Next.js guide](/observability/frontend/how-to/setup-nextjs.md).

**Simple alternative:** Use `https://telemetry.nav.no/collect` directly in your production code. The URL is stable and all production clusters use the same endpoint.