Menu
Find naturally clean, coherent video segments worth keeping (selection over repair)
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Generate karaoke-style word-level timestamps by aligning script text to audio using Qwen3-ForcedAligner + jieba for Chinese word segmentation. Use when the user says 'align captions', 'karaoke timestamps', 'word timestamps', 'caption alignment', 'sync text to audio'.
Analyze raw video content using Gemini to identify speakers, topics, key moments, and potential clip opportunities
Audio processing utilities - noise reduction, normalization, enhancement
Extract a video segment using FFmpeg with precise start/end times
Text-guided audio source separation using SAM-Audio via mlx-audio
ASR with ~30ms timestamp precision using Qwen3-ASR + ForcedAligner
| name | find-golden-segments |
| description | Find naturally clean, coherent video segments worth keeping (selection over repair) |
Use this skill to identify the best moments in raw footage - segments that are already clean and don't need repair.
Selection over Repair: Instead of trying to fix broken footage by cutting out fillers, find the moments that are naturally good.
python skills/find-golden-segments/find_golden.py <video_path>
# Examples
python skills/find-golden-segments/find_golden.py video.MOV
python skills/find-golden-segments/find_golden.py video.MOV --min-duration 15
{
"golden_segments": [
{
"start": "02:15",
"end": "02:38",
"duration_sec": 23,
"score": 9,
"speaker": "Speaker Name",
"topic": "Brief topic",
"quote_preview": "First few words...",
"quality_notes": "Why this segment is good"
}
],
"summary": {
"total_duration": 140,
"golden_duration": 45,
"segments_found": 3
}
}
Only segments scoring 7/10 or higher are included. Lower-quality sections are skipped entirely.