# Demo AI Face Service

This LWA demonstrates the complete browser-to-Gateway face workflow:

1. `@cutos/ai-face-detector` captures one qualified face locally.
2. `@cutos/gw-client` connects with the CUTOS device credentials obtained from `@cutos/core`.
3. `@cutos/ai-face-service` creates the selected collection and invokes registration, lookup, listing, search, comparison, removal, and clearing.

During SDK development, the demo references `file:../sdk`. After `@cutos/ai-face-service` is published, replace it with `^4.0.0` and regenerate `package-lock.json`.

## Run

```bash
npm install
cutos lwa build
cutos lwa upload --upload-mode oss
cutos lwa publish --device <device-id>
```

The target CUTOS Runtime must provide device information including Gateway credentials (`id`, `gwi`, `token`, and `gwUrl`).