podcast-audio-processing

What problem does it solve?

The post-production workflow for podcasts—including audio conversion, transcription, and creating chapter markers—is a multi-step, time-consuming, and often manual process. This skill automates the entire audio processing pipeline, saving hours of repetitive work.

Core Features & Use Cases

• Audio Conversion: Automatically converts .m4a audio files to .mp3 (128kbps) for optimal size and compatibility.
• Local Transcription: Utilizes OpenAI Whisper (base model) for fast, private, and accurate transcription, generating a full JSON transcript.
• Automated Chaptering: Analyzes the generated transcript to identify natural topic transitions and creates 10-15 descriptive chapter markers in both FFmpeg and Podcasting 2.0 formats.
• Chapter Embedding: Embeds the generated chapter metadata directly into the final .mp3 file, enhancing listener experience in supported podcast apps.
• Use Case: Take a raw audio file from NotebookLM, automatically convert it, transcribe it, generate intelligent chapter markers based on content, and embed them, producing a fully podcast-ready .mp3 and transcript for publishing.

Quick Start

Use the podcast-audio-processing skill to process the audio file 'Original_Audio.m4a' for the episode at 'podcast/episodes/2025-12-01-topic-slug' with episode slug '2025-12-01-topic-slug'.