ow-start-transcribe

原文涵盖的主题

• /ow-start-transcribe
• Steps
• Rules

Part of ow-tools, the code add-on for the ow- skills. It runs code on this machine: ffmpeg, whisper.cpp and pyannote, through the ow-tools command. Nothing is uploaded. The speaker model runs with network access off. 1. Setup. If the argument is --setup, or the user asks what is needed, run ow-tools doctor and show its output as it is. Every fix is a command for the user to run in their own terminal. Never run brew, uv sync, a download or a keychain command yourself. Stop here. 2. Find the recording. It must be in a folder named audio inside the workstream, e.g. opchain-work/intake/ /audio/ .m4a. Naming the folder is enough when it holds one recording. If the recording is somewhere else, ask the user to move it there. Don't move it yourself. 3. Transcribe. Run ow-tools transcribe --json. Add --speakers N when the head count is known (the intake's intake.yaml may list attendees). Add --out for a second recording in the same workstream, e.g. plans/ /workshops/W1/. If the workstream's handling level is set and opchain-work/profile.md has a bannertext, pass it verbatim with --banner " ". A 20-minute call takes a few minutes; say so before you start. 4. Handle the result by exit code. 0

这里只展示节选，请以完整原文和当前要求为准。