ReadyTrace NAS

• 

  IMAGE

ReadyTrace NAS — evidence-first release readiness for creative teams

What if your NAS could tell you not just which files exist, but which deliverables are actually ready for release—and show the evidence?

Creative studios and educational publishers maintain master specifications alongside hundreds of lesson PDFs, illustrations, narration files, and video exports. A missing voiceover, wrong file type, or two versions with the same filename can derail a handoff at the last minute.

ReadyTrace NAS is an original Uliya orchestration Skill for UGREEN AI NAS. It audits only the folders a user authorizes, extracts expected assets from a nominated master brief, and compares them with actual files to produce a transparent, source-cited readiness report.

Core workflow

1. Authorize a bounded folder, master guide, and audit depth; do not inspect unrelated NAS content.
2. Discover canonical paths and file metadata through Uliya's registered file expert.
3. Extract requirements, source sections, item IDs, and expected formats through the document expert.
4. Reconcile deliverables using deterministic ID/type matching. Optional image or audio checks require user authorization and supported experts.
5. Report found, missing, ambiguous, conflicting, and unverified requirements, with precise file paths, coverage, and recommended next actions.
6. Save only with approval. The user reviews the full report and its destination first. The Skill never renames, moves, overwrites, or deletes original materials.

What makes it different

Typical file search finds assets; generic AI summaries describe them. ReadyTrace links a release requirement to local evidence and explicitly refuses to invent a match. It distinguishes duplicate basenames, checks required file types, preserves source provenance, and issues a human-review gate instead of pretending that an inventory scan constitutes editorial approval.

Technical design

• Target: UGREEN UGOS / Uliya; Markdown-only SKILL.md with directly linked references.
• Expert orchestration: file-agent (ug-filesystem/ug-filemgr), document-agent (ug-document), optional photo-agent (ug-photo), knowledge-agent (ug-knowledge).
• Safety: read-only source workflow, limited scan scope, platform permission gates for any saved report, no direct NAS shell access, no bundled executables or external API calls.
• Error states: completed, failed, cancelled, waiting_user, and waiting_approval; missing permissions or uncertain data yield an honest partial report.

Development and validation

This Skill and its core workflow were authored for the 2026 UGREEN Hackathon, beginning October 28, 2026. The Round 1 ZIP contains the full orchestration contract, evidence rules, synthetic scenario, permissions and fallback handling, and a finalist test plan. It passed local format/structure checks. Not yet tested on real UGREEN NAS, and not claimed to have passed the organizer's downloadable static checker.

If selected as a finalist, we will verify expert signatures and run a 12-item complete synthetic release, a partially missing/duplicated release, and an edge case with inaccessible or changing files. We will measure false matches, report traceability, and approval integrity.

AI / third-party disclosure

OpenAI ChatGPT (GPT-6) was used for concept development, drafting, review, and test planning. Python standard library packaged the ZIP; ReportLab (BSD license, reportlab.com/opensource) generated the separate PDF. No third-party runtime libraries, scripts, cloud services, private datasets, or externally hosted models are included in the Skill. The intended runtime uses only Uliya's registered platform capabilities under actual permissions.

Track: Automate Your Workflow

Round 1 deliverables: Proposal PDF (10 answers, under 500 words) and pre-test Skill ZIP. No NAS hardware or demonstration video is required in Round 1.

Built With

• ai-agents
• chatgpt
• markdown
• ugos
• ugreen-ai-nas
• uliya