本文へスキップ
TakoAPI
スキル一覧に戻る

audio-conductor

wells1137 作

Web & Frontend Development
—まだ評価がありません
あなたの評価

概要

Intelligently dispatches requests to the appropriate audio generation model (Music, Sound Effects, or TTS)

説明

このスキルでできること

Routes audio generation requests to the appropriate specialized model based on prompt analysis. Accepts a single unified request and dispatches to music, sound effects, or text-to-speech generation internally. Depends on the elevenlabs-mcp-server skill being loaded in the agent environment.

Eliminates the need to manually route requests to separate audio models by centralizing classification and delegation behind a single call.

使いどころ

  • Generating background music for a video without picking a model
  • Creating a door creak or ambient noise sound effect
  • Producing a voice-over from a written script
  • Building audio pipelines without hardcoding model selection logic
  • Letting an agent handle mixed audio tasks through one interface

ワークフロー例

入力

User asks: Generate a tense cinematic swell for my trailer intro

エージェント
  1. 1

    Receives the prompt and optional params like duration_ms or instrumental

  2. 2

    Analyzes the prompt for keywords indicating music, sfx, or tts category

  3. 3

    Classifies the request as music based on mood and instrument references

  4. 4

    Delegates to the music-generator sub-skill with a Composition Plan

  5. 5

    Formats the returned audio file into a standard AudioOutput object

出力

AudioOutput with audio_url, duration_ms, audio_type: music, and model_used metadata

要件

  • ElevenLabs API key (required by elevenlabs-mcp-server)

インストール

clawhub install wells1137/audio-conductor

CLIコマンドを実行するか、Claude Code、Cursor、Codexなどお使いのコーディングエージェントにプロンプトを貼り付けてください。

リンク