Cut a 6-hour training recording into highlights
Web page · https://agentusecases.ai/use-case/cut-a-6-hour-training-recording-into-highlights
Done with Claude Code. A training team used Claude Code with video-editing skills to transcribe a 6-hour live recording locally, find pauses and irrelevant parts, and build a Kdenlive project for manual touch-ups.
First seen by Oriel on .
AI summary
A training team used Claude Code to locally transcribe a six-hour recording, remove pauses and irrelevant sections, and create a Kdenlive project for manual editing, reducing the work from about three days to just over one.
- Feng Yu combined the videocut-skills and cli-anything-kdenlive skills without using the paid Volcano Engine transcription API.
- The pipeline used faster-whisper, applied rules to cut verbal slips and pauses, and generated a Kdenlive project file; a CLI bug needed fixing before it opened.
- Because the machine could not run faster-whisper large-v3, the team used the small model at int8 and transcribed chunks in parallel, reducing timing accuracy.
- The rough cut still contained many verbal slips and redundancies and produced too many clips, so Feng polished it by hand.
#claude code #faster-whisper #kdenlive #video editing
Generated by gpt-5.6-sol. Check the original for details.
Aliases
| Source | External ID | URL |
|---|---|---|
| Agent Use Cases | https://agentusecases.ai/use-case/cut-a-6-hour-training-recording-into-highlights | https://agentusecases.ai/use-case/cut-a-6-hour-training-recording-into-highlights Read discussion |