Inspiration

Terabytes of high-quality family memories and content creation footage often end up sitting unused on local storage due to the tedious, time-consuming nature of manual video editing. NAS-CineCraft-AI was created to bridge this gap by bringing autonomous, agentic video orchestrator capabilities directly to local UGREEN AI NAS storage—ensuring total media privacy without cloud bandwidth limits.

What It Does

NAS-CineCraft-AI functions as a local, autonomous media orchestrator:

  • Local Media Discovery: Scans target folders using UGOS Files and Albums capabilities.
  • Smart Timeline Generation: Indexes video clips and photos by timestamp, location, and visual quality to form a chronological highlight script.
  • Privacy-First Processing: Runs entirely on-device with zero external data retention or cloud uploads.
  • User Safety Confirmation: Always requests explicit user approval before writing render output files to storage.

How We Built It

We structured NAS-CineCraft-AI around the official Uliya Skill Development Standard:

  1. Media Ingestion Layer: Interfaces with local NAS file paths and UGOS indexing tools.
  2. Orchestration Layer: Builds logical scene transitions, timestamps, and story timelines.
  3. Execution Layer: Leverages local CLI capability execution (FFmpeg composition) to render finished highlight videos upon user confirmation.

Challenges We Faced

  • Designing an agentic workflow that runs efficiently on local NAS compute without triggering memory overhead during heavy multi-clip processing.
  • Strictly adhering to UGOS security boundaries, ensuring zero missing parameters, no placeholder code (TODOs), and enforcing explicit user confirmation gates for all file write operations.

What We Learned

  • How to maximize local edge capabilities of UGREEN AI NAS to process high-resolution media locally.
  • How to design production-ready, standardized Markdown agent skills under the Uliya Skill Development Standard.

What's Next for NAS-CineCraft-AI

  • Expanding multi-track audio indexing for automated background score generation.
  • Adding interactive preview capabilities within the UGOS user dashboard upon advancing to the Finalist round!

Built With

Share this project:

Updates

Submission history