AgentSkillsCN

audiobook-studio

制作一部内容连贯、叙述统一、且可导出的多章节有声书。

SKILL.md
--- frontmatter
name: audiobook-studio
description: Produce a full multi-chapter audiobook with consistent narration and exportable files.
allowed-tools: list_dir, read_file, upload_file, tts_async_create, tts_async_query, retrieve_file, download_file, list_files, delete_file

You are running the Audiobook Studio skill.

Goal

  • •Produce a multi-chapter audiobook with consistent narration and clear output organization.

Ask for

  • •Folder or list of chapter files (preferred: .txt).
  • •Voice preference or ask to browse voices.
  • •Output format (mp3 or wav), and any pacing/volume notes.

Workflow

  1. •Discover chapters (list_dir) and confirm ordering.
  2. •For each chapter:
    • •Upload the chapter text file (purpose: t2a_async_input).
    • •Create a tts_async_create task with the chosen voice settings.
  3. •Poll tasks in batches with tts_async_query.
  4. •Download or retrieve each completed file.
  5. •Return a final list of audio file paths with chapter order.

Notes

  • •Keep voice settings consistent across all chapters unless the user requests per-character voices.
  • •If the user wants a sample before running the full batch, generate one chapter first and confirm.