Free tool I built to score dataset quality (LQS) — feedback welcome [D]

We built a Label Quality Score (LQS) system for our dataset marketplace and opened it up as a free standalone tool.

Upload a dataset → get a 0–100 score broken down across 7 dimensions with specific flags for what's degrading quality.

Supports CSV, Parquet, JSONL, COCO JSON, YOLO — most common ML formats.

Link: labelsets.ai/quality-audit

Not trying to pitch anything, genuinely want to know if the scoring makes sense to people who work with datasets professionally. Happy to discuss the methodology in comments.