clean-audio — Claude Code skill: audio cleanup and installation
The clean-audio Claude Code skill cleans voice audio in video, reducing hum, background noise, hiss, and harsh sibilance. It activates on audio-cleanup requests or after a rough cut.
MIT ★ 541 updated 2026-11-09
npx skills add dotslashsimran/simcraft --skill clean-audio -a claude-codeWhat clean-audio does
The skill measures the audio, then chooses a cleanup level that preserves speech clarity and lip sync. It checks a short sample first; a full run reports the speech-to-background gap, final loudness, and detected issues such as hum or clipping.
Clean the raw clips before editing: a rough cut removes the pauses used to identify noise. The skill needs Python from the rough-cut virtual environment and Whisper. If that environment does not exist yet, create it with sh ~/simcraft/skills/video/rough-cut/setup.sh. Its filters cannot remove voices or music. Original files remain untouched.
Set denoising with --strength, loudness with --target, or skip tone shaping with --no-eq. The listed target is -14 LUFS for YouTube, Instagram, and TikTok, and -16 LUFS for podcasts.
What it helps with
- remove hum, hiss, and background noise from speech
- even out voice levels in a video
- reduce harsh sibilance and muddy or harsh tone
- clean audio without losing spoken words
- process raw clips before making a rough cut
Example prompts for the agent
- Clean the audio in
interview.mp4and save a video with the processed track. - Remove hum and background noise from the source clips in this workspace.
- Set the podcast audio target to -16 LUFS.
How to install clean-audio
1Skills CLI (Vercel)
npx skills add dotslashsimran/simcraft --skill clean-audio -a claude-code2Manually: put the files in place
~/.claude/skills/clean-audio/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/clean-audio/
1Skills CLI (Vercel)
npx skills add dotslashsimran/simcraft --skill clean-audio -a codex2Manually: put the files in place
.agents/skills/clean-audio/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add dotslashsimran/simcraft --skill clean-audio -a cursor2Manually: put the files in place
.cursor/skills/clean-audio/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install dotslashsimran/simcraft skills/video/clean-audio2Manually: put the files in place
.github/skills/clean-audio/SKILL.md1Skills CLI (Vercel)
npx skills add dotslashsimran/simcraft --skill clean-audio -a gemini-cli2Platform CLI
gemini skills install https://github.com/dotslashsimran/simcraft --path skills/video/clean-audioManually: put the files in place
.devin/skills/clean-audio/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about clean-audio
How do I install clean-audio in Claude Code?
Run npx skills add dotslashsimran/simcraft --skill clean-audio -a claude-code. The skill file is at ~/.claude/skills/clean-audio/SKILL.md.
What does clean-audio do?
It reduces noise, hum, and harsh sounds in video voice tracks, and adjusts loudness. A clarity check prevents the cleanup from removing spoken words.
What does clean-audio need to run?
It requires Python from the rough-cut virtual environment and Whisper. Create the environment with sh ~/simcraft/skills/video/rough-cut/setup.sh if it is not already present.
Source text
SKILL.md · MIT
---
name: clean-audio
description: Clean up voice audio in a video: rumble, electrical hum, background noise and hiss, muddy or harsh tone, sharp S sounds, uneven volume, and loudness for YouTube/Instagram/TikTok. Measures the audio first, picks the strongest cleanup that doesn't make the words less clear, and keeps lip sync exact. Use when the user says "clean up the audio", "fix the sound", "remove background noise", "remove hum", "my audio is too quiet/loud", "make it sound better", or after a rough cut.
---
# Clean audio
Run with the rough-cut skill's Python: `PY=~/simcraft/skills/video/rough-cut/.venv/bin/python` (run `sh ~/simcraft/skills/video/rough-cut/setup.sh` once if it doesn't exist). It shares rough-cut's helpers and uses Whisper for the clarity check.
## Steps
1. **Listen test first.** Make a 10-second before/after of the busiest talking moment. Both are set to the same loudness, so the comparison is about quality, not volume:
```
$PY scripts/clean_audio.py <video> --sample
```
Open both wav files for the user (`open <before> <after>`) and give them the report.
2. **Apply.** There are two ways to run it:
- In a rough-cut workspace, before rendering. It cleans each raw clip into `audio/<clip>.clean.wav`, and rough-cut's `render.py` uses those automatically:
```
$PY scripts/clean_audio.py <workspace>
```
Then render and verify as usual. The report is saved to `audio/cleanup.md`.
- On a single video. It writes `<name>_clean.<ext>` next to the original, with the video copied untouched:
```
$PY scripts/clean_audio.py <video>
```
3. **Report** in plain words:
- the speech-vs-background gap before and after,
- the final loudness,
- what was found (hum, clipping),
- which strength the clarity check settled on.
If the report says leftover noise is voices or music, tell the user that these filters can't remove it, and suggest an AI voice isolator.
## Options
- `--strength strong|medium|light` sets the strongest denoise to try. The default is medium, and the clarity check steps it down on its own.
- `--target -14` sets the loudness in LUFS. -14 is right for YouTube, Instagram and TikTok; use -16 for podcasts.
- `--no-eq` skips the tone shaping, if the user likes their voice as it is.
## Rules
- **Clean the raw clips before cutting, not after.** The noise is learned from a pause, and the rough cut removes the pauses.
- Never touch the original files.
- Never trade clarity for quiet. The clarity check enforces this: no lost words, and confidence within 0.015 of the original.
- Don't push audio to digital silence between words. A little room sound is kept on purpose.
What each step does, why it's built this way, and the test numbers: [reference.md](reference.md).
© dotslashsimran · MIT · dotslashsimran/simcraft