Audio Prompt Reverse logo

Audio Prompt Reverse

Community
JGRFW
audio-prompt-reverse

数字人提示词生成。根据一张或多张人物演唱图和一段歌曲音频,使用带时间码的真实歌词,生成可直接粘贴到 11 号 MiniMax H3 数字人工作流的六段式提示词。支持单人和男女对唱;使用 wav2vec2 男女声分类模型结合音色聚类识别声部,音高算法仅作兜底。对唱使用交替单人镜头,每个歌词段画面中只有对应歌手,另一人完全在画外。

Overview

PublisherJGRFW
Repositorycomfyui-AICG3D-SKILLS
Skill nameaudio-prompt-reverse
Stars
119
Forks
22
Bundled files
4
LicenseGPL-3.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 4 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by JGRFW on GitHub. Read the source before you install it.

Installation

Install the Audio Prompt Reverse AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/JGRFW/comfyui-AICG3D-SKILLS/tree/main/skills/audio-prompt-reverse
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/JGRFW/comfyui-AICG3D-SKILLS.git /tmp/comfyui-AICG3D-SKILLS
mkdir -p .claude/skills
cp -r /tmp/comfyui-AICG3D-SKILLS/skills/audio-prompt-reverse .claude/skills/audio-prompt-reverse
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Audio Prompt Reverse in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Audio Prompt Reverse on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Audio Prompt Reverse is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

数字人提示词生成

把「人物演唱图 + 歌曲音频 + 带时间码歌词」转换成可直接用于 MiniMax H3 数字人模式的六段式提示词。单人歌曲使用一个数字人;男女对唱必须使用两个独立人物和共同场景。

输入

  • <Picture N>:人物演唱图。单人模式默认 <Picture 1>;双人对唱根据用户描述分别绑定男子、女子和共同场景。
  • 歌曲音频:由 11 号工作流的 AICG3D 资源库接入,作为数字人模式全局锁定的最终音轨。
  • 歌词时间线:由插件使用 Whisper 识别,或由用户通过“台词时间线”输入校正。歌词和时间码是最终定稿。

输出规则

  1. 只输出以下六个字段,顺序固定:
text
subject_definitions:
summary:
retention_analysis:
detailed_description:
overall_soundscape:
non_diegetic_music:
  1. 单人模式把人物写成 <Subject 1>、歌手 (S1)。双人对唱必须分别定义男声 <Subject 1>(S1)、女声 <Subject 2>(S2)和共同场景 <Subject 3>;如果用户使用其他图片编号,则按实际编号绑定。
  2. 不要在最终提示词中使用 <Audio N> 标签。11 号数字人节点会自动删除这类标签;音频关系改用“锁定歌曲音轨”说明。
  3. retention_analysis 必须同时包含每张人物图、场景图的 fully_preserved,以及 锁定歌曲音轨: fully_copy - ...。
  4. detailed_description 必须从 [Shot 1] 开始,并在同一字段中写入真实演唱时间表。每个时间段使用:
text
[Shot 2] 在 00:03.560 处,句首硬切到女声单人镜头。
00:03.560-00:07.120, <Subject 2> (S2) sings:
<d>[Mandarin singing] 真实歌词原文</d>

[Shot 1] 不带切点时间码;从 [Shot 2] 起,每个镜号后写 在 MM:SS.mmm 处,时间码必须严格递增。 6. 歌词必须逐字使用识别结果,保留原语言、标点和顺序;不得翻译、改写、补词、拆句或合并。 7. 每个歌词时间段前写清当前歌手的口型、下颌、嘴唇、呼吸、眼神、头部和肩颈如何跟随原歌曲同步。声部优先使用 wav2vec2 男女声分类模型,并用音色聚类做交叉确认;模型不可用或结果不确定时才回退音高。确认男女两个声部后,才使用交替单人镜头:当前歌手是画面中有且只有的一个人,另一名歌手完全在画外。 自动声部判定不可靠时禁止猜成男女交替;会把状态标记为“不确定”,并保留完整时间线供手动校正“台词时间线”中的 S1/S2。 8. 每个镜头只承担一个主要机位动作。按段落选择不同镜头语法:前景横移揭示、低机位侧向 Truck、三分之四近景 Dolly、约 30° 受控小弧线、手持跟随、Crane/Pedestal 下降或结尾 Crane Up + Dolly Out。对唱镜头始终保持单人入画。 9. 三个以上歌词段时,至少使用三种真正不同的镜头,不要全片只靠 Zoom、Push In 或固定机位。切换可以借鉴 MV 的硬切、动作匹配切、擦镜切和重音直切,但不要使用溶解或随机堆叠术语。 10. 切镜只放在下一句歌词起唱、换气或明确重音处,不能切在咬字中间。歌手切换时,上一歌手在切点前闭唇并完全离开画面,新镜头只出现下一位歌手。起唱句必须保留当前歌手的完整嘴部和可读面部,共享场景与屏幕方向保持连续。 11. 镜头可以更有创意,但人物身份、发型、服装、配饰和场景空间关系不能漂移。不要新增原素材中没有的人物、道具、地点、文字或光源。 12. overall_soundscape 和 non_diegetic_music 写 N/A,因为完整歌曲音轨已由数字人模式锁定。 13. 不输出分析、摘要、Markdown 围栏、操作说明或第六段之后的任何文字。

结果必须可以直接复制粘贴到 11 号工作流的 MiniMax H3 节点,节点模式选择“数字人”。

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Audio Prompt Reverse AI skill do?

数字人提示词生成。根据一张或多张人物演唱图和一段歌曲音频,使用带时间码的真实歌词,生成可直接粘贴到 11 号 MiniMax H3 数字人工作流的六段式提示词。支持单人和男女对唱;使用 wav2vec2 男女声分类模型结合音色聚类识别声部,音高算法仅作兜底。对唱使用交替单人镜头,每个歌词段画面中只有对应歌手,另一人完全在画外。

Why use Audio Prompt Reverse on TypingMind?

Because you install it once and use it with any model. Audio Prompt Reverse is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Audio Prompt Reverse in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/JGRFW/comfyui-AICG3D-SKILLS/tree/main/skills/audio-prompt-reverse. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Audio Prompt Reverse?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Audio Prompt Reverse?

As many as you like. As long as a model supports skills, you can use Audio Prompt Reverse with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Audio Prompt Reverse AI skill free?

Yes. It is published on GitHub by JGRFW under the GPL-3.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇