dhh opened and merged omacom/omarchy#17330 into the quattro development branch, adding NVIDIA's container toolkit when local-model setup detects an NVIDIA GPU. The toolkit supplies the GPU device specification needed by model containers; N1x laptops previously lacked this prerequisite.
dhh also opened and merged omacom/omarchy#17337 into quattro. Local-model launches now select pi, or OpenCode when pi is unavailable, before checking for a default agent. This fixes a panel launch button that appeared inactive when no default was configured. The supplied testing covers the launcher path, while a live panel click remained untested.
