- Add text-to-audio pipeline using stable-audio-open-1.0
- Implement route handler and worker methods
- Add Docker configuration
- Update OpenAPI specs and Go bindings
- Update port mappings and dependencies

The pipeline accepts text prompts and generates audio using
stabilityai/stable-audio-open-1.0. Features include:
- Adjustable audio duration (1-30 seconds)
- Configurable inference steps and guidance scale
- Optional negative prompt support
- Reproducible generation via seed parameter