Inspiration
People store important information across personal documents, bills, receipts, assignments, and project notes. Yet finding the right information does not always answer a more important question: What should I do next?
We created NextWise AI to bridge the gap between stored information and actionable priorities, while keeping recommendations grounded in evidence.
What it does
NextWise AI — Know What Matters. Know What's Next.
NextWise AI is an evidence-based personal priority assistant designed for UGREEN AI NAS.
Its proposed workflow helps users:
- Discover obligations and deadlines in authorized NAS documents.
- Reconcile obligations with matching evidence, such as bills and payment receipts.
- Identify completed, unresolved, overdue, and conflicting items.
- Prioritize the next action based on documented consequences, deadlines, and dependencies.
- Review the source evidence supporting each recommendation.
Missing payment or completion evidence is never treated as proof that an obligation remains unfinished.
How we built it
NextWise AI was designed using the official UGREEN UGOS Skill Creation Tool v0.7.3 and the Uliya Skill Development Standard.
The Skill uses Markdown-defined orchestration and the supported expert-agent architecture, including:
ug-filesystemthroughfile-agentfor authorized file discovery.ug-documentthroughdocument-agentfor information extraction and reconciliation.
The workflow includes evidence-based status classification, an explainable priority policy, claim re-checking, and explicit failure handling.
The initial implementation is read-only and does not include custom executables or external data services.
Challenges we ran into
The main challenge was designing meaningful reasoning without relying on unsupported scripts, direct NAS access, or custom runtimes.
Other challenges included handling ambiguous dates, avoiding false completion claims, matching documents correctly, and distinguishing supported evidence from assumptions.
We addressed these issues through strict evidence requirements, defined status categories, bounded document scope, and transparent uncertainty reporting.
Accomplishments that we're proud of
- Created an original evidence-based priority workflow for personal NAS documents.
- Designed cross-document reconciliation and explainable next-action selection.
- Added explicit handling for missing evidence, conflicting information, and permission failures.
- Passed the official Uliya static Skill checker with 0 errors and 0 warnings.
Current validation status: Static checks passed. The Skill has not yet been tested on a real UGREEN AI NAS.
What we learned
Reliable AI assistants need more than information retrieval. They must distinguish evidence from inference, explain recommendations, and recognize uncertainty.
We also learned that effective AI NAS Skills depend on careful orchestration of supported capabilities rather than adding unnecessary software components.
What's next for NextWise AI
If selected as a finalist, we plan to deploy and test NextWise AI on the assigned UGREEN AI NAS environment.
Our validation priorities are deadline extraction, evidence reconciliation, next-action ranking, source accuracy, repeated-run behavior, and failure handling.
Longer term, NextWise AI could support broader personal knowledge and productivity workflows, subject to available platform capabilities and user permissions.
NextWise AI turns stored information into evidence-based next actions.
Built With
- agents
- ai
- markdown
- nas
- ugos
- ugreen
- uliya
