skills.rest
marioplumbeer
marioplumbeerCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

plumbeer-recording-transcriber

Transcribe audio recordings to timestamped text using mlx-whisper on Apple Silicon Macs.

Updated Mar 22, 2026
Search Tags:#transcription#whisper#macos#timestamps#speech to text#audio to text#mlx-whisper
One-click install
npx skills add https://github.com/marioplumbeer/skills --skill plumbeer-recording-transcriber-marioplumbeer
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: plumbeer-recording-transcriber
Source: https://github.com/marioplumbeer/skills/tree/main/.agents/skills/plumbeer-recording-transcriber
Command: npx skills add https://github.com/marioplumbeer/skills --skill plumbeer-recording-transcriber-marioplumbeer

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires mlx-whisper, ffmpeg, and includes scripts (resource) and references (resource) components.

What problem does it solve? Turning spoken audio recordings into written transcripts usually requires manual typing or cloud services. This Skill converts local audio files into text transcripts with per-segment [MM:SS] timestamps, then files the transcript and source audio together under a topic-language-timestamp filename. ## Core Features & Use Cases - Local transcription with mlx-whisper: Runs the whisper-large-v3-turbo model on Apple Silicon via a persistent Python virtualenv, with ffmpeg for audio decoding. - Guided naming workflow: Confirms the detected language with the user, proposes five topic options based on the transcript content, and names outputs as topic_language_timestamp. - Organized output: Saves the transcript and a copy of the source audio side by side, defaulting to ~/Recordings, with a visible step-by-step progress checklist. - Use Case: You record a voice memo about a project idea. Point the Skill at the file, confirm the detected language, pick a topic like "product roadmap", and receive product-roadmap_en_20250115_143022.txt plus a matching audio copy. ## Quick Start Transcribe the most recent audio file in ~/Recordings and save the transcript with timestamps next to it.

Frequently Asked Questions about plumbeer-recording-transcriber

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

FAQPage Schema
How do I transcribe an audio file to text on a Mac?▼

Point the Skill at an audio file or let it pick the most recent one in ~/Recordings. It runs mlx-whisper's large-v3-turbo model locally, writes a transcript with [MM:SS] timestamps per segment, and saves it next to a copy of the source audio.

What is mlx-whisper used for in audio transcription?▼

mlx-whisper runs OpenAI's whisper speech-to-text models on Apple Silicon via the MLX framework. This Skill uses the mlx-community/whisper-large-v3-turbo checkpoint, which is tuned for Apple Silicon throughput, and prints the detected language as a final LANGUAGE: code line.

Does this transcription skill work on Intel Macs or Linux?▼

No. mlx-whisper only runs on Apple Silicon Macs, and the Skill checks uname for Darwin and arm64 before installing anything. If the platform check fails, it stops and tells the user rather than falling back to a slower engine.

Why does my transcript end with repeated phrases like Thank you?▼

Repeated short phrases at the tail are a known whisper hallucination artifact caused by trailing silence in the recording. The Skill flags this in the delivery summary rather than silently editing the repeated text out of the transcript.

Can I get a plain text transcript without timestamps?▼

Yes. Timestamps are on by default, producing [MM:SS] prefixed lines per segment, but you can opt out in Step 1. The Skill then passes --no-timestamps to the script, which outputs one flat text block instead.

How are transcript output files named?▼

Filenames follow the pattern topic-slug_language-code_timestamp, where the timestamp is the source audio file's own modification time formatted as YYYYMMDD_HHMMSS. The transcript gets a .txt extension and the copied audio keeps its original extension.

Related Skills

Explore More Personal & Entrepreneur Skills→
LuisvelMarketerLuisvelMarketer
1

openai-whisper

Transcribe audio files locally using the Whisper CLI.

Community
Intermediate
Lucien-1127Lucien-1127

youtube-transcript

Extracts YouTube video transcripts and organizes them into timestamped structured notes.

Community
Basic
lidge-junlidge-jun
5

transcribe

Transcribe audio files into text with speaker diarization using OpenAI's Whisper models.

Community
Intermediate
rustykuntzrustykuntz
1

mlx-whisper

Transcribe and translate local audio files using MLX Whisper on Apple Silicon.

Community
Intermediate