What problem does it solve? Long meeting recordings are hard to review and extract value from. This Skill turns audio or video recordings into a structured written digest covering discussion topics, decisions, and open questions, without requiring a paid transcription API. ## Core Features & Use Cases - Audio Preparation: Converts any media file to mono 16kHz MP3 with ffmpeg and splits it into 20-minute overlapping chunks sized for LLM context windows. - Gemini Gem Workflow: Prepares chunk files and placeholder digest files so the user can process them through a Gemini Gem, then pastes results back. - Digest Assembly: Merges per-chunk summaries into a single digest.md, removing overlap duplicates and organizing content into sections like decisions and unanswered questions. - Use Case: After a two-hour team sync recorded as an MP4, run the Skill to chunk the audio, process it through Gemini, and receive a clean structured summary with decisions and open questions. ## Quick Start Ask the agent to digitize a meeting by saying "оцифруй встречу" and providing the path to your recording file.
