What it does

Local Media Transcription processes local audio and video files to produce readable transcriptions and structured meeting deliverables. It supports common media formats (MP4, M4A, MP3, WAV, WEBM) and workflows for Japanese and multilingual meetings, offering raw transcripts, speaker-separated outputs (when diarization is available), meeting minutes, cleaned customer-facing summaries, action-item lists, and PowerPoint-friendly bullet points. The skill documents tool checks and decision points (ffmpeg/ffprobe availability, whisper CLI availability) and emphasizes quality checks on transcription outputs.

When to use it

Use this skill when you need to transcribe recorded meetings, training sessions, or local screencast videos without sending media to remote services. It's suitable for internal reviews, producing post-meeting minutes, or preparing customer-facing summaries where some human verification is expected. If diarization tools are unavailable, the skill falls back to a plain transcription and clearly documents limitations.

What's included

Scripts: none bundled; the workflow references CLI tools (ffmpeg, whisper) and gives sample commands for running transcription.
References: transcription-workflow.md with decision points for diarization, tool checks, and output formatting.
Instructions: stepwise workflow for input validation, media probing, transcription command, quality sniffing, optional diarization, and downstream formatting for minutes/action items.