transcribe
Audio Transcription Pipeline
VAD-first, evidence-backed Whisper transcription running locally on Apple Silicon. Produces high-quality transcripts with zero hallucinations, correct domain terminology, and LLM-polished output.
Architecture: Audio → ffmpeg (16kHz WAV) → Silero-VAD (speech segmentation) → MLX Whisper → Dictionary replacement → Claude LLM correction → Output.
Prerequisites
- macOS with Apple Silicon (M1/M2/M3/M4) — required for MLX
- Python 3.9+ — Xcode Python at
/Library/Developer/CommandLineTools/usr/bin/python3or Homebrew Python - ffmpeg — for audio preprocessing (
brew install ffmpeg) - Anthropic API key — for LLM correction step (optional but recommended)
Setup
Detection Order
Check if the pipeline is already installed:
More from simonstrumse/vibelabs-skills
bunny-net
Use Bunny.net for CDN, image hosting, video hosting, edge storage, image optimization, AI image generation, edge scripting, container hosting, DNS, and security (WAF/DDoS). Trigger when user mentions bunny, cdn, image hosting, video hosting, media storage, edge storage, image optimization, video streaming, pull zone, storage zone, WebP conversion, AVIF, lazy loading, DRM, video encoding, edge functions, serverless edge, container hosting, WAF, DDoS protection, bot detection, or content delivery.
25klarsprak
Norwegian plain language (klarspråk) guidelines from Språkrådet. Use when writing Norwegian website copy, marketing text, UI labels, documentation, or any user-facing Norwegian content. Triggers on Norwegian text, norsk tekst, klarspråk, markedsføring, nettsidetekst, brukergrensesnitt, or when reviewing/improving Norwegian copy quality.
5fiken
|
3company-research
Company research using Exa search. Finds company info, competitors, news, tweets, financials, LinkedIn profiles, builds company lists.
3soft-glass-ui
Design system for creating premium glass UI using iOS 26 native Liquid Glass APIs (.glassEffect modifier) or fallback gradient-based approaches for older iOS versions. Use when building SwiftUI iOS apps with glassmorphism, premium UI, translucent cards, or modern Apple-style interfaces. Triggers on requests for liquid glass, glassmorphism, iOS 26 design, premium UI, or translucent interfaces.
3ios-development
Comprehensive iOS app development skill with Liquid Glass design, modern Swift patterns (@Observable, SwiftData), and App Store readiness. Self-updates by checking Apple Developer News for new iOS versions. Use for any iOS/SwiftUI project.
3