npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

@adatechnology/audio-transcription-mcp

v0.1.0

Published

MCP server exposing local Whisper transcription (whisper.cpp) as a tool — no API key, no upload

Readme

@adatechnology/audio-transcription-mcp

Servidor MCP que expõe transcrição de áudio por Whisper rodando na própria máquina: sem chave de API, sem upload, sem custo por minuto. É um invólucro fino sobre o engine local do @adatechnology/audio-transcription-provider.

Pré-requisitos

Três coisas fora do processo: o binário whisper-cli (whisper.cpp), o ffmpeg e um modelo ggml no disco.

brew install whisper-cpp ffmpeg          # macOS
curl -L -o ~/models/ggml-large-v3-turbo.bin \
  https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-large-v3-turbo.bin

Use large-v3-turbo (~1.5GB). Modelos menores trocam palavras e entram em loop de repetição em pt-BR — o motivo está detalhado no README do provider.

Registrar no Claude Code

claude mcp add ada-transcription -- \
  bunx @adatechnology/audio-transcription-mcp

Com as variáveis de ambiente:

| Variável | Obrigatória | Padrão | Para quê | |---|---|---|---| | WHISPER_MODEL_PATH | ✅ | — | caminho do modelo ggml | | WHISPER_BINARY_PATH | | whisper-cli | binário do whisper.cpp | | FFMPEG_PATH | | ffmpeg | conversão para WAV 16kHz mono | | WHISPER_THREADS | | do whisper.cpp | threads na inferência | | TRANSCRIPTION_LANGUAGE | | pt | idioma padrão | | TRANSCRIPTION_ALLOWED_ROOT | | — | limita a leitura a uma pasta |

Sobre TRANSCRIPTION_ALLOWED_ROOT: quem escolhe o caminho do arquivo é o modelo, não uma pessoa digitando. Sem essa variável o servidor lê qualquer arquivo que o usuário do processo consiga ler. Aponte para a pasta de áudios e o problema deixa de existir.

Ferramentas

transcribe_audio{ filePath, languageHint? } → texto. Caminho absoluto, na máquina que roda o servidor. Aceita m4a, mp3, ogg/opus, wav, flac, webm, amr e aac.

check_transcription_setup — sem argumentos. Diz qual dos três pré-requisitos falta. Use quando transcribe_audio falhar, antes de investigar o áudio.

Medição

Áudio real de reunião gravado por celular (7min38s, m4a), via MCP de ponta a ponta, Apple M3 Pro com Metal: 23,3s, 4.964 caracteres, sem loop de repetição. Em CPU de container espere bem mais lento.

Testes

bun test