Minutes Record

Records meetings and calls by capturing audio from your mic, transcribing it locally with whisper.cpp or parakeet.cpp, and saving timestamped markdown files to ~/meetings/. Works for in-person conversations on MacBook's built-in mic or system audio from Zoom and Meet if you route it through BlackHole. The live transcript streams to a JSONL file while recording so you can follow along mid-meeting. Everything stays on your machine, no cloud transcription. The tiny model is fast but sloppy, small is the sweet spot for real meetings, and large gets you the best accuracy if you can wait. Transcripts are saved with 0600 permissions since they contain sensitive content. Takes a few minutes of setup to download a speech model but then it just works.

GitHub stars: 1.3K
last updated: Jun 11, 2026

View on GitHub

Install

$ npx -y skills add silverstein/minutes --skill minutes-record --agent claude-code

Installs into .claude/skills of the current project.

Files

/minutes-record

Record audio from the microphone, transcribe it locally with whisper.cpp, and save as searchable markdown.

How it works

Recording is a two-step process — start and stop. Between those two commands, audio is captured continuously from the default input device.

Start recording:

minutes record
# Or with a title:
minutes record --title "Weekly standup with Alex"

The process runs in the foreground. It captures audio from whatever input device is active — the built-in MacBook mic for in-person conversations, or a BlackHole virtual audio device for system audio (Zoom, Meet, Teams calls).

Stop recording:

minutes stop

This sends a signal to the recording process, which then:

Stops audio capture
Transcribes the audio locally via whisper.cpp (no cloud, no data leaves the machine)
Saves the transcript as a markdown file in ~/meetings/
Prints the output path and word count as JSON

Live transcript during recording:

skill.md on GitHub