Dead Air Removal
When to Use
- A CapCut draft folder already exists
- Source media is already present on the timeline
- The user wants silence trimming without manual CapCut UI editing
Preflight Checks
- Confirm
<draft-folder>/draft_content.jsonexists - Confirm
<draft-folder>/Timelines/project.jsonexists - Confirm source media paths still resolve
- Refuse to patch when validation fails or nearly all content would be removed
Default Command
Run python -m scripts.workflow --draft-folder "<draft-folder>" from the skills/dead-air-removal/ directory (repo checkout) or the skill install directory.
Gentler Silence Removal (less aggressive)
Run python -m scripts.workflow --draft-folder "<draft-folder>" --silence-threshold-db -35 --min-silence-ms 250 --lead-padding-ms 120 --tail-padding-ms 180 --max-gap-ms 120 --min-keep-ms 300 from the same directory.
Outputs
dead_air_report.jsondead_air_report.md- timestamped backup files next to the patched draft
Requirements
- Run this from the skill directory that contains
scripts/workflow.py - Python available in the local environment
ffmpegavailable onPATH, or pass--ffmpeg-bin "<path-to-ffmpeg>"- Agent can read files and run shell commands