メニュー
AI-assisted annotation review — uses YOLO inference + Claude vision to auto-approve, correct, or flag images for human review in CVAT.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Set up autonomous training monitoring — creates cron jobs to track long-running training, auto-continue pipeline when training completes.
Audit YOLO dataset quality — class distribution, annotation quality, image stats, and improvement suggestions.
Orchestrate the full active learning loop: train, analyze, push to CVAT, wait for review, pull, merge, retrain.
Analyze YOLO training runs — compares to baseline/best, checks per-class regression, analyzes training dynamics and tune convergence, writes actionable recommendations.
Run autonomous YOLO training experiments — reads training-plan.md, assesses bottlenecks, acts strategically, and delegates HP optimization to model.tune().
Profile YOLO model inference speed, FPS, and size across image sizes and export formats.
| name | review-annotations |
| description | AI-assisted annotation review — uses YOLO inference + Claude vision to auto-approve, correct, or flag images for human review in CVAT. |
Multi-pass annotation review: YOLO precision + Claude intelligence.
STATUS: EXPERIMENTAL
Read yolo-project.yaml for classes, model path. Get review folder from user.
Create: review_output/approved/, needs_review/, rejected/
For each image:
python scripts/multi_pass_annotate.py --image <path> --model <model> --conf 0.15python scripts/draw_annotations.pyWrite review_output/review_report.md with counts, corrections, common issues.