Shoot Cull Assistant

Inspiration

Event, wedding and portrait photographers often come home with 500 to 3,000 frames. The first cull (blurred, blinking and duplicate shots) takes hours, and most culling tools live on one laptop or need the photos uploaded somewhere.

What it does

Shoot Cull Assistant is a Uliya Skill that works on the shoot folder where it already sits on the NAS. It reads EXIF, groups bursts, asks the vision model to judge each frame against a rubric for the shoot type, and shows counts, the least certain calls and any unreadable files. The photographer fixes a few labels and accepts; only then are NAS file tags and a cull report written. Reruns replace old cull tags instead of stacking new ones, and nothing is ever deleted.

How we built it

SKILL.md workflow plus references (scoring rubric, report template, test plan, disclosures), using ug-filesystem, ug-photo (meta, vision-vlm), ug-filemgr tags, ug-knowledge and file-workflow. Uliya delegates to file-agent, photo-agent and knowledge-agent, and every write waits for platform approval. The package passes the Uliya Skill Kit 0.7.3 static checker with zero errors.

What's next

If selected as a Finalist: tune the prompt and thresholds on synthetic shoots, measure time per frame and accuracy on the NAS, and confirm RAW handling and where vision-vlm runs.

Built With

• file-workflow
• ug-filemgr
• ug-filesystem
• ug-knowledge
• ug-photo
• ugreen-ai-nas
• uliya