Agent-Skills.md

Agent Skills: OpenAI Whisper API (音频转录)

Transcribe audio via OpenAI Audio Transcriptions API (Whisper).

UncategorizedID: Jst-Well-Dan/Skill-Box/openai-whisper-api

Repository

Jst-Well-Dan/Skill-Box
Jst-Well-Dan
74

Install this agent skill to your local

pnpm dlx add-skill https://github.com/Jst-Well-Dan/Skill-Box/tree/HEAD/i18n/skills/content-pipeline/openai-whisper-api

Skill Files

Browse the full folder contents for openai-whisper-api.

Download Skill

Loading file tree…

i18n/skills/content-pipeline/openai-whisper-api/SKILL.md

Skill Metadata

Name
openai-whisper-api
Description
通过 OpenAI 音频转录 API (Whisper) 转录语音。支持多种音频格式，可提供提示以优化转录效果。

OpenAI Whisper API (音频转录)

通过 OpenAI 的 /v1/audio/transcriptions 端点转录音频文件。

快速开始

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

默认设置：

  • 模型 (Model): whisper-1
  • 输出 (Output): <输入文件名>.txt

常用参数

{baseDir}/scripts/transcribe.sh /path/to/audio.ogg --model whisper-1 --out /tmp/transcript.txt
{baseDir}/scripts/transcribe.sh /path/to/audio.m4a --language zh
{baseDir}/scripts/transcribe.sh /path/to/audio.m4a --prompt "演讲者姓名：彼得，丹尼尔"
{baseDir}/scripts/transcribe.sh /path/to/audio.m4a --json --out /tmp/transcript.json

API 密钥

设置 OPENAI_API_KEY 环境变量，或在 ~/.openclaw/openclaw.json 中配置：

{
  skills: {
    "openai-whisper-api": {
      apiKey: "在此输入您的 OpenAI 密钥",
    },
  },
}