skills.rest
Kaleb-Rupe
Kaleb-RupeCOMMUNITY
Explore all skills by this author →
𝕏@MightieMagsGitHub

AudioEditor

Transcribes audio with Whisper and removes filler words, stutters, and dead air via ffmpeg edits.

Updated Jul 10, 2026
Search Tags:#transcription#whisper#ffmpeg#podcast#audio editing#filler words#cleanvoice
One-click install
npx skills add https://github.com/Kaleb-Rupe/aurora --skill audioeditor-kaleb-rupe
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: AudioEditor
Source: https://github.com/Kaleb-Rupe/aurora/tree/main/claude/skills/Utilities/AudioEditor
Command: npx skills add https://github.com/Kaleb-Rupe/aurora --skill audioeditor-kaleb-rupe

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Manually cleaning podcast or voice recordings of filler words, stutters, false starts, and long pauses is tedious and time-consuming. This Skill automates the entire editing pipeline from transcription to final cut. ## Core Features & Use Cases - Word-Level Transcription: Uses insanely-fast-whisper (MPS accelerated) or standard Whisper to generate timestamped transcripts. - LLM Edit Classification: Claude classifies segments as filler, stutter, false start, edit marker, or dead air, distinguishing rhetorical emphasis from mistakes. - Automated Cutting: ffmpeg applies cuts with 40ms qsin crossfades, with optional Cleanvoice cloud polish for mouth sounds and loudness normalization. - Use Case: Clean a raw podcast recording by running the full pipeline to produce an edited MP3 with ums, stutters, and dead air removed, optionally previewing proposed edits first. ## Quick Start Ask the assistant to clean up the audio on your podcast file and remove filler words and dead air.

Frequently Asked Questions about AudioEditor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I remove filler words from a podcast recording automatically?▼

Run the pipeline with your audio file to transcribe it with Whisper, classify filler words and stutters using Claude, and cut them with ffmpeg crossfades. Use the --aggressive flag for tighter detection of single filler words and shorter pauses.

How can I preview audio edits before applying them?▼

Run the pipeline with the --preview flag to transcribe and analyze the audio without modifying it. It displays each proposed edit with timestamps, type, reason, and context, then you rerun without the flag to apply them.

What audio formats does this editing pipeline support?▼

The edit step supports MP3, WAV, FLAC, and M4A/AAC files, preserving the original codec and bitrate where possible. Output defaults to <filename>_edited.<ext> in the same directory as the input.

What API keys are required for automated audio cleaning?▼

ANTHROPIC_API_KEY is required for the LLM analysis step and is loaded from ~/.config/aurora/.env. CLEANVOICE_API_KEY is only needed if you use the optional --polish flag for cloud-based mouth sound removal and loudness normalization.

Why did transcription fail or produce no output?▼

Transcription requires either insanely-fast-whisper or the standard whisper CLI installed; the script falls back to standard whisper if the fast variant fails. Install one with pip, and note long files can take several minutes even on MPS.

Related Skills

Explore More Content & Communication Skills→
ZJU-REALZJU-REAL
195

lark-minutes

Search Feishu Minutes recordings and retrieve metadata, media, and transcripts via lark-cli.

Official
Advanced
Marky-TeamMarky-Team
3

schedule-posts

Schedule and publish social media posts with captions and media via the Marky API.

Official
Advanced
skills-ilskills-il
6

hebrew-podcast-postproduction

Generate a Hebrew podcast post-production bundle from a transcript or SRT file.

Official
Advanced
the-open-agentthe-open-agent
5.5k

nano-pdf

Edit specific PDF pages using natural-language instructions via the nano-pdf CLI.

Official
Intermediate