Overview
The UGREEN Hackathon 2026 on AI NAS is an online developer challenge hosted on Devpost. It invites participants to build AI Skills that run on UGREEN AI NAS devices. UGREEN AI NAS is powered by UGOS, a Linux-based operating system that combines local storage, computing, and AI capabilities in one platform. Using the UGOS Skill Creation Tool and their preferred AI assistant, participants can create custom Skills that use supported NAS capabilities to organize files, find information, and automate everyday tasks. The event runs from October 22, 2026, with the submission deadline on November 15, 2026 at 11:45 pm PST.
Who can participate
Participation is open to individuals who are above the legal age of majority in their country of residence. Companies and professional organizations are not eligible. A list of excluded countries and territories applies, including China, India, Russia, and Ukraine. No UGREEN NAS is required to enter Round 1.
Tracks
Participants choose one track that best fits their Skill:
- Automate Your Workflow: Use authorized NAS files and data, together with the file, document, knowledge, and system capabilities provided, to organize content, extract information, generate summaries, and streamline repetitive tasks.
- Bring Memories to Life: Work with authorized photos, videos, and documents to organize or search content, extract text, generate summaries, or create albums.
- Unleash Your Creativity: Help photographers, editors, designers, and other content creators search, tag, select, organize, and review media stored on the NAS.
- Open Innovation: Combine NAS, AI, media, file, and knowledge capabilities to build an original Skill that addresses a real user need.
Format and rounds
Round 1 (open to all) requires participants to download the UGREEN Skill Developer Package and submit two items by the deadline: a project proposal PDF of no more than 1 MB, covering ten items in no more than 500 English words in total, and a Pre-test Skill Package in ZIP format. A NAS and a demo video are not required at this stage.
Round 2 (Finalists only) is limited to up to 30 Finalists, who will be invited by email on November 21. Each Finalist receives one assigned day of remote access to a UGREEN AI NAS iDX6011 Pro, between November 22 and November 28, to install, test, and polish the Skill. Finalists must submit a final Skill package and a public YouTube demo video of no more than 3 minutes, with English narration or complete English subtitles, by November 30, 2026 at 11:45 pm PT.
Prizes
The total cash prize pool is $46,000:
- Grand Prize: $20,000 (1 winner)
- Second Place: $10,000 (1 winner)
- Third Place: $6,000 (1 winner)
- Most Creative: $5,000 each (2 winners), based on the highest scores for Innovation & Differentiation
Judging criteria
Submissions are evaluated by a panel of qualified judges on four criteria:
- User Value: whether the Skill solves a clear user problem and delivers a clear workflow, useful feedback, and a real result.
- Innovation & Differentiation: whether the Skill offers a new function, interaction, or workflow different from common AI tools.
- Technical Quality: whether core and repeat-use cases work, errors and edge cases are handled, and data, permissions, credentials, and dependencies are managed well.
- AI NAS Fit & Validation: whether UGREEN AI NAS features add real value and whether the Skill uses NAS resources well.
Getting started
Participants are encouraged to learn about UGREEN AI NAS, join the UGREEN Community, and download the UGREEN Skill Developer Package, which contains the required tools and technical documentation. Useful skills include Python or JavaScript/TypeScript, Linux and Docker, AI agent and tool design, and file or media processing.
Submission requirements
The Skill package must follow the Uliya Skill Development Standard. A SKILL.md file is required, with frontmatter containing a lowercase, hyphenated name of 2 to 64 characters and a description of up to 1,024 characters. The ZIP file and its uncompressed contents must each be no larger than 10 MB, with a maximum of 1 MB per file and 100 files. Third-party components, open source libraries, and AI tools used must be disclosed with their sources, licenses, and principal uses. Configuration examples must use placeholders, not real credentials or personal data.
