# Video-Processing-Expert

> Domain-specific skills for navigating and implementing the Dancebox AI video editing pipeline.

- Author: kunaldhongade
- Repository: kunaldhongade/dancebox-backend
- Version: 20260206170843
- Stars: 0
- Forks: 0
- Last Updated: 2026-02-06
- Source: https://github.com/kunaldhongade/dancebox-backend
- Web: https://mule.run/skillshub/@@kunaldhongade/dancebox-backend~Video-Processing-Expert:20260206170843

---

---
name: Video-Processing-Expert
description: Domain-specific skills for navigating and implementing the Dancebox AI video editing pipeline.
---

# Video-Processing-Expert

Use this skill to implement or modify the core AI video editing logic.

## 1. Domain Context

- **Raw Input**: Talking-head videos ≤ 2 minutes.
- **Output**: 9:16 vertical shorts.
- **Logic**: Smart cuts, B-roll insertion, and automated captioning.

## 2. File Organization

- **Job Flow**: Starts in `src/queues/`, processed in `src/workers/`.
- **Logic**: Business rules for video rendering live in `src/services/video/`.
- **Integrations**: AI services (Speech-to-Text) live in `src/services/ai/`.

## 3. Implementation Rules

- Always use `BullMQ` for long-running processing tasks.
- Ensure all video assets are correctly enqueued and status is updated in Prisma.
- Adhere to the `9:16` aspect ratio requirement without variation.