mcpbeat Sign in

Openai Whisper Agent Skill

by coco-research

Speech-to-text transcription via OpenAI Whisper. Supports two modes — Local CLI (no API key, runs on-device) and Cloud API (fast, scalable, requires OPENAI_API_KEY). Use when the user needs to transcribe audio files, translate speech, or convert audio to text.

967 tokens
context cost
the whole folder, loaded on every use
1
files
instructions only
0
copies elsewhere
how many repositories repackaged it
196
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository 
npx skills add https://github.com/coco-research/coco --skill openai-whisper

The instruction itself

13 sections, as written by the author

OpenAI Whisper — Speech-to-Text

Transcribe audio files using OpenAI's Whisper model. Two modes available depending on your needs:

| Mode | Latency | Cost | Privacy | Setup |

|------|---------|------|---------|-------|

| Local CLI | Slower (on-device GPU/CPU) | Free | Audio never leaves machine | Install whisper binary |

| Cloud API | Fast | Per-minute pricing | Audio sent to OpenAI | OPENAI_API_KEY required |

Mode 1: Local CLI

Run Whisper locally with no API key required. Models download to ~/.cache/whisper on first run.

Quick Start

whisper /path/audio.mp3 --model medium --output_format txt --output_dir .

Common Commands

# Transcribe to text file
whisper /path/audio.mp3 --model medium --output_format txt --output_dir .

# Transcribe with translation to English
whisper /path/audio.m4a --task translate --output_format srt

# Transcribe with specific language
whisper /path/audio.wav --model large --language en --output_format json

Model Selection

| Model | Speed | Accuracy | VRAM |

|-------|-------|----------|------|

| tiny | Fastest | Lowest | ~1 GB |

| base | Fast | Low | ~1 GB |

| small | Medium | Good | ~2 GB |

| medium | Slow | Better | ~5 GB |

| large | Slowest | Best | ~10 GB |

| turbo | Fast | Good (default) | ~6 GB |

Output Formats

  • txt — Plain text transcript
  • srt — SubRip subtitle format with timestamps
  • vtt — WebVTT subtitle format
  • json — Detailed JSON with word-level timestamps
  • tsv — Tab-separated values

Notes

  • --model defaults to turbo on most installs
  • Use smaller models for speed, larger for accuracy
  • GPU acceleration used automatically when available

Mode 2: Cloud API

Transcribe via OpenAI's /v1/audio/transcriptions endpoint. Faster for large batches, no local GPU needed.

Quick Start

{baseDir}/scripts/transcribe.sh /path/to/audio.m4a

Defaults:

  • Model: whisper-1
  • Output: <input>.txt

Common Commands

# Basic transcription
{baseDir}/scripts/transcribe.sh /path/to/audio.m4a

# Specify model and output
{baseDir}/scripts/transcribe.sh /path/to/audio.ogg --model whisper-1 --out /tmp/transcript.txt

# With language hint
{baseDir}/scripts/transcribe.sh /path/to/audio.m4a --language en

# With speaker name hints (improves accuracy)
{baseDir}/scripts/transcribe.sh /path/to/audio.m4a --prompt "Speaker names: Peter, Daniel"

# JSON output with timestamps
{baseDir}/scripts/transcribe.sh /path/to/audio.m4a --json --out /tmp/transcript.json

Raw curl Example

curl https://api.openai.com/v1/audio/transcriptions \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "Content-Type: multipart/form-data" \
  -F file="@/path/to/audio.m4a" \
  -F model="whisper-1" \
  -F response_format="text"

API Key Setup

Set OPENAI_API_KEY environment variable, or configure in ~/.clawdbot/clawdbot.json:

{
  skills: {
    "openai-whisper-api": {
      apiKey: "OPENAI_KEY_HERE"
    }
  }
}

Choosing Between Modes

| Consideration | Local CLI | Cloud API |

|---------------|-----------|-----------|

| Privacy-sensitive audio | Best | Audio sent to OpenAI |

| Large batch processing | Slow without GPU | Fast and parallel |

| Offline usage | Works offline | Requires internet |

| Cost | Free (hardware cost) | Per-minute pricing |

| Setup complexity | Install binary + models | API key only |

| Audio format support | Most formats | Most formats |

Other skills for the same job

different authors, same section of the catalogue
Dhdna Profiler
by K-Dense-AI
×1

Extract cognitive patterns and thinking fingerprints from any text. Use this skill when the user wants to analyze how someone thinks, understand cognitive style, profile writing or speech patterns, compare thinking styles between people, asks "what's my thinking style", "analyze how this person reasons", "cognitive profile", "thinking pattern", "DHDNA", "digital DNA", or wants to understand the mind behind any text. Also trigger when the user provides text and wants deeper insight into the author's reasoning patterns, decision-making style, or cognitive signature.

4k tokens
Gsap
by openai
vendor ×1

GSAP animation reference for HyperFrames. Covers gsap.to(), from(), fromTo(), easing, stagger, defaults, timelines (gsap.timeline(), position parameter, labels, nesting, playback), and performance (transforms, will-change, quickTo). Use when writing GSAP animations in HyperFrames compositions.

5k tokens scripts
Image Assistant
by ComeOnOliver
×1

配图助手 - 把文章/模块内容转成统一风格、少字高可读的 16:9 信息图提示词；先定“需要几张图+每张讲什么”，再压缩文案与隐喻，最后输出可直接复制的生图提示词并迭代。

15k tokens scripts zh
Youtube Clipper
by nexu-io

| YouTube clip generation and editing with automated workflows — pull source video, slice highlights, add captions, and export.

298 tokens
Interactivity Best Practices
by remotion-dev
vendor

Best practices for writing Remotion animations that stay intuitive for agents and editable in Remotion Studio Visual Mode.

146 tokens
Youtube Transcript
by browser-act

YouTube transcript extraction and content reformatting: given a YouTube video URL, opens the video's transcript panel, extracts all timestamped segments, and transforms the raw transcript into summaries, chapter outlines, Twitter/X threads, blog posts, or notable quotes. Use when the user shares a YouTube URL or video link, asks to summarize a video, get a transcript, extract content from a YouTube video, get YouTube captions, extract YouTube captions, download YouTube captions, transcribe YouTube video, YouTube video to text, make a thread from YouTube, YouTube to blog post, YouTube to article, pull transcript from YouTube, YouTube content extraction, convert YouTube to text, video to transcript. Also applies when user wants to reformat any YouTube video content into structured output (chapters, threads, blog articles, key quotes).

3k tokens scripts
Ecommerce Full Pipeline
by anbeime

跨境电商全链路自动化工具。集成1688采集、智能清洗、多平台上架（微信小店/Shopify/TikTok）、推广方案（关键词/竞品分析/广告文案）、短视频创作（MoviePy竖屏视频）、一键代发、爆品挖掘（趋势聚合+6维评分）、闲鱼二手选品捡漏（品牌识别/虚标过滤/捡漏评分/价格监控）、全自动流水线（挖掘→采集→清洗→上架→推广→视频）。

2k tokens zh
Historical Interview Scripts
by anbeime

生成历史名人现代访谈短视频文案，通过古今反差与网络热梗的爆笑结合，创作具有传播力的虚构趣味内容

50k tokens scripts zh

How to use it

Copy the folder

Take coco-research/openai-whisper from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.