video-upscalerlisted

## Summary The **Video Upscaler** skill provides professional-grade video quality enhancement by leveraging a powerful, multi-model backend. It intelligently selects the best AI model (Topaz, SeedVR2, etc.) based on the user-defined profile to achieve optimal results, transforming low-resolution or noisy footage into crisp, cinematic-quality video. This skill abstracts away the complexity of choosing and configuring different AI upscaling models. Instead of dealing with dozens of technical parameters, the user simply chooses a high-level goal, and the skill handles the rest. ## Features - **Multi-Model Backend**: Dynamically routes requests to the best model for the job (Topaz, SeedVR2, etc.) via a unified API. - **Profile-Based Enhancement**: Offers a range of pre-configured profiles for common use cases, from standard 2x upscaling to 4K cinematic conversion and 60 FPS frame boosting. - **Asynchronous by Design**: Handles long-running video processing jobs without blocking the agent. - **Simple Interface**: Requires only a video URL and a profile name to start. ## How It Works The skill operates in a simple, two-step asynchronous workflow: 1. **Submit Job**: The agent calls the `/upscale` endpoint with a video URL and a profile name. The service validates the request, selects the appropriate AI model, and submits the job to the `fal.ai` backend. It immediately returns a `task_id`. 2. **Poll for Status**: The agent uses the `task_id` to periodically call the `/statu