-
MediaMorph AI checks your UGREEN NAS media library before playback and flags readiness issues with safe, explainable next steps.
Inspiration
I built MediaMorph AI around a simple problem: a NAS media library can look organized, but users often discover issues only when they actually try to watch something.
A title may not appear correctly, subtitles may be missing, or the library may be inconsistently organized. I wanted to move that discovery earlier and create a pre-flight readiness check for a private NAS media library.
My goal is simple: before someone presses Play, MediaMorph should help answer:
Is this media ready, does something need attention, or do I not have enough evidence to know?
What it does
MediaMorph AI is a local-first Uliya Skill for UGREEN AI NAS.
I select a media folder or title and ask MediaMorph to check its playback readiness.
The Skill coordinates supported UGREEN capabilities and produces one of three explainable results:
- Ready — no relevant issue was found from the available signals.
- Needs Attention — MediaMorph found an actionable issue such as missing subtitles, discovery problems, or inconsistent organization.
- Unknown — the available capabilities do not provide enough evidence to make a reliable decision.
MediaMorph does not silently modify the user's library. It explains the evidence, prioritizes the issues, and recommends safe next steps. Any supported organization change requires explicit user confirmation.
How I built it
I designed MediaMorph as a coordinator Skill that delegates work only to registered UGREEN experts.
The core flow is:
Files → Video/Music → Knowledge → Local Uliya synthesis
I use Files to inspect the user-authorized media scope, Video/Music to search media and subtitle-related information, and Knowledge to retrieve relevant official UGREEN guidance.
I then combine those results using deterministic readiness rules where possible and use the local Uliya model to explain and prioritize the findings.
I created the SKILL.md package and supporting Markdown references using the supplied Uliya Skill Development Kit. The package passes the supplied static validation checks.
Challenges we ran into
The biggest challenge was creating something useful without pretending that undocumented UGREEN capabilities exist.
The public capability set does not guarantee codec-level inspection, transcoding, direct playback testing, unrestricted system access, or automatic NAS configuration.
Instead of hiding that limitation, I designed MediaMorph around evidence confidence.
If the available signals are insufficient, the Skill returns Unknown rather than inventing a diagnosis.
I believe that makes the experience safer and more trustworthy.
Accomplishments that we're proud of
I am proud that MediaMorph is designed around the actual UGREEN Skill architecture rather than as a generic AI concept.
I have:
- Defined the media-readiness problem and user workflow.
- Mapped the solution to documented UGREEN expert capabilities.
- Built the coordinator
SKILL.md. - Added explicit failure and uncertainty handling.
- Added user-confirmation safeguards for supported changes.
- Kept the core workflow local-first.
- Passed the supplied Uliya static validation checks.
- Avoided unsupported scripts, executables, credentials, or direct NAS access.
Most importantly, MediaMorph does not try to replace the media player. It acts as a readiness layer before playback.
What we learned
I learned that building a trustworthy AI assistant is not only about adding more automation.
A useful NAS Skill also needs to know when not to act.
For MediaMorph, the distinction between Ready, Needs Attention, and Unknown became an important design principle.
I also learned that the strongest UGREEN integration is not a standalone chatbot. The value comes from coordinating existing NAS capabilities and turning fragmented signals into one understandable decision.
What's next for MediaMorph AI
If selected as a finalist, I will validate MediaMorph on the provided UGREEN AI NAS.
My next steps are to:
- Test real Files, Video/Music, and Knowledge expert responses.
- Validate subtitle and media-discovery behavior.
- Test empty, inaccessible, and poorly organized libraries.
- Measure false-positive and false-negative readiness classifications.
- Refine deterministic readiness rules using real NAS behavior.
- Improve evidence explanations and severity ranking.
- Explore deeper media inspection only if UGREEN exposes an officially supported capability for it.
My goal is to evolve MediaMorph from a Skill specification into a reliable, privacy-first media readiness guardian for UGREEN AI NAS.
Log in or sign up for Devpost to join the conversation.