Agent Skill
2/7/2026package-release
Assemble all processed audio files and metadata into final release structure ready for distribution. Use after audio post-processing is complete or when the user requests to package the final audiobook release.
X
xiaxianlin
0GitHub Stars
1Views
npx skills add xiaxianlin/video-book
SKILL.md
| Name | package-release |
| Description | Assemble all processed audio files and metadata into final release structure ready for distribution. Use after audio post-processing is complete or when the user requests to package the final audiobook release. |
name: package-release description: Assemble all processed audio files and metadata into final release structure ready for distribution. Use after audio post-processing is complete or when the user requests to package the final audiobook release.
打包发布 Skill
目的
将所有处理后的音频文件和元数据打包为最终发布结构,准备分发。
输入
build/05_post/chapters/*.mp3- 章节音频文件build/01_extracted/chapters.json- 章节元数据build/02_casting/voice_mapping.json- 音色分配build/03_segmentation/segment_manifest.json- 片段统计build/05_post/processing_log.json- 处理统计
输出
release/audio/chapters/*.mp3- 最终章节音频文件release/audio/full_book.mp3- 可选的全书音频(如果请求)release/meta.json- 完整项目元数据release/chapters.json- 章节信息release/README.md- 发布文档
执行指令
当此 skill 被调用时:
-
复制章节音频文件
cp build/05_post/chapters/*.mp3 release/audio/chapters/ -
生成全书音频(可选) 如果请求,连接所有章节:
# 创建文件列表 ls release/audio/chapters/*.mp3 | sort > filelist.txt # 连接 ffmpeg -f concat -safe 0 -i filelist.txt \ -c copy \ release/audio/full_book.mp3 -
创建 meta.json 编译综合元数据:
{ "project": { "name": "audiobook_project", "version": "1.0.0", "generation_date": "2026-01-26T14:00:00Z", "pipeline_version": "v1.0" }, "source": { "input_file": "input/source.md", "total_chapters": 10, "total_word_count": 25000 }, "audio": { "total_duration_seconds": 11325, "total_duration_formatted": "3h 8m 45s", "format": "MP3", "bitrate": "192kbps", "channels": "mono", "sample_rate": "44100Hz" }, "production": { "tts_provider": "openai", "tts_model": "gpt-4o-mini-tts", "total_segments": 150, "target_loudness_lufs": -18, "characters_count": 5 }, "characters": [ { "character_id": "char_001", "name": "张三", "voice": "onyx", "total_lines": 150 } ] } -
创建 chapters.json 播放器的章节级信息:
{ "chapters": [ { "chapter_number": 1, "title": "第一章标题", "audio_file": "audio/chapters/ch_001.mp3", "duration_seconds": 1125, "duration_formatted": "18m 45s", "word_count": 2500, "file_size_mb": 2.8 } ] } -
生成 README.md 创建发布文档:
# 有声书发布 ## 项目信息 - **标题**:[小说标题] - **生成日期**:2026-01-26 - **总时长**:3h 8m 45s - **格式**:MP3, 192kbps, Mono ## 内容 - 10 个章节 - 多角色旁白 - 专业音频处理 ## 音频文件 - `audio/chapters/` - 单独的章节文件 - `audio/full_book.mp3` - 完整有声书(可选) ## 技术规格 - TTS 提供商:OpenAI gpt-4o-mini-tts - 响度:-18 LUFS - 真峰值:-1.0 dBTP - 片段时长:~75 秒 ## 角色与音色 [角色和分配音色列表] ## 使用方法 按顺序播放章节文件,或使用 full_book.mp3 连续播放。 -
验证发布包
- 所有章节 MP3 存在
- 元数据文件为有效 JSON
- 文件大小合理
- 音频文件可播放
- README 完整
-
生成发布清单
{ "release_date": "2026-01-26T14:00:00Z", "files": [ { "path": "audio/chapters/ch_001.mp3", "size_bytes": 2936832, "md5": "abc123...", "duration_seconds": 1125 } ], "total_size_mb": 28.5, "checksum": "sha256:def456..." } -
报告结果
- 发布中的文件总数
- 总包大小
- 发布目录结构
- 验证状态
- 分发就绪状态
验证
- 所有预期文件存在
- JSON 文件有效且完整
- 音频文件可播放
- 元数据准确
- README 信息丰富
可选增强
- 生成封面艺术占位符
- 创建 M4B 有声书格式
- 添加章节标记
- 生成播客分发的 RSS feed
- 创建分发存档(ZIP/TAR)
Skills Info
Original Name:package-releaseAuthor:xiaxianlin
Download