Music Video logo

Music Video

Organization
NoizAI
music-video

Build a music video from an existing song and lyrics, either as p5.js/p5.brush animation or by assembling local images and clips. Use when the user mentions a music video, MV, lyric video, visualizer, storyboard, 音乐视频, 歌词动画, or MV合成. Do not use it to compose the song, generate sound effects, or call an image or video generation API.

Overview

PublisherNoizAI
Repositoryskills
Skill namemusic-video
Stars
526
Forks
80
Bundled files
3
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.

  • 3 bundled files

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

  • Open source

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

Installation

Install the Music Video 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/NoizAI/skills/tree/main/skills/music-video
  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/NoizAI/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/skills/music-video .claude/skills/music-video
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Music Video 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 Music Video 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 Music Video 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.

music-video

Turn one finished song into a 24fps music video. The song is the master audio. Scene intervals cover it exactly once.

Triggers

  • music video / MV / lyric video / visualizer / storyboard
  • 音乐视频 / MV合成 / 歌词动画 / 歌曲动画

Capabilities

  • code: initialize a pinned p5.js and p5.brush project, then animate the song in JavaScript.
  • hybrid: inspect local character images, scene stills, and video clips, then assemble them. No image or video generation API is called.
  • Local beat, onset, low-frequency, and loudness analysis for cut points and animation cues.
  • Frame-exact assembly with the original song as the only soundtrack.

Inputs

Required:

  • song: local audio (.mp3, .wav, .m4a, .flac, .ogg, .aac)
  • brief: the visual idea, audience, and any required references
  • mode: code or hybrid

Optional:

  • lyrics: text or .srt. Required before storyboarding when the vocal timing matters.
  • assets: directory of stills and clips for hybrid
  • aspect: 16:9 (default), 9:16, 1:1, or 4:3

If the user wants a song created first, use text-to-music, then return here with the chosen file.

Outputs

  • <project>/output/final.mp4
  • docs/CREATIVE_BRIEF.md, docs/STORYBOARD.md, docs/TASKS.md, and docs/PROGRESS.md
  • audio/analysis.json, and docs/inspect.json when assets exist

Workflow

Run python3 skills/music-video/scripts/mv.py from the repository root. If this skill is installed elsewhere, use the scripts/mv.py beside this file. Create the project outside the skill directory.

  1. Choose code when the video is drawn in JavaScript. Choose hybrid when the user supplies stills or clips. Ask once when neither is clear.
  2. Initialize an isolated project:
bash
python3 skills/music-video/scripts/mv.py init \
  --project ./mv-project --song ./song.mp3 --lyrics ./lyrics.txt --mode code

python3 skills/music-video/scripts/mv.py init \
  --project ./mv-project --song ./song.mp3 --lyrics ./lyrics.srt \
  --assets ./clips --mode hybrid --aspect 16:9
  1. Analyze the song. Treat tempo changes as approximate and confirm section boundaries from the audio:
bash
python3 skills/music-video/scripts/mv.py analyze --project ./mv-project
  1. Get lyric timing. If the user did not supply it, use speech-to-text only when NOIZ_API_KEY or ~/.config/noiz/api_key already exists. Never print the key. If neither exists, ask for lyrics or timestamps and do not upload the song.
  2. Read references/opus-5.5-prompt.md and follow it. Write the creative brief and storyboard before scene code or assembly. Review the opening and main peak before building the other chapters.
  3. In code mode, read <project>/animation/ANIMATION_GUIDE.md before drawing. Set PROJECT.bpm and PROJECT.offset from audio/analysis.json. The pinned template canvas is 1920×1080; change studio.html and the renderer window before using another aspect. Check motion with the template renderer:
bash
node render.mjs --sheet=1,2,3 --out=../output/sheets/pilot.jpg
node render.mjs --strip=START:END --out=../output/sheets/pilot-motion.jpg
node render.mjs --clip --range=START:END --out=../output/pilots/opening.mp4

Run those inside <project>/animation. Pass --chrome or set CHROME_PATH when Chrome is not in a standard location. A contact sheet only screens layout. A --strip, or playback of the opening, peak, and ending, is the motion review. 7. In hybrid mode, inspect local media and stop if a required clip is missing. Do not generate a replacement.

bash
python3 skills/music-video/scripts/mv.py inspect --project ./mv-project
  1. Write timeline.json using references/timeline.schema.json. Intervals are inclusive at start_frame and exclusive at end_frame.
  2. Assemble and probe the result:
bash
python3 skills/music-video/scripts/mv.py assemble --project ./mv-project

Timeline

json
{
  "fps": 24,
  "width": 1920,
  "height": 1080,
  "audio": "audio/source.mp3",
  "scenes": [
    {
      "id": "opening",
      "start_frame": 0,
      "end_frame": 240,
      "source": "output/pilots/opening.mp4",
      "source_start": 0,
      "fit": "cover",
      "transition": "fade",
      "fade_frames": 6
    }
  ],
  "overlays": []
}

fade stays inside its own scene. Do not overlap scenes to create a crossfade.

Completion

Finish only when every item is true:

  • [0, ceil(duration × 24)) is covered exactly once, with no gap or overlap.
  • final.mp4 has a video stream and the complete original song.
  • The opening, main peak, and ending were reviewed in motion.
  • The last story read has time to land.
  • docs/PROGRESS.md has no open blocker.

Security

init --mode code clones JohnHeibel/ClaudeAnimationBase at commit 0ac8bf2b31942376cb6b8c4074715595d512acd2 into the project. That template is MIT licensed. Analysis, inspection, rendering, and assembly stay local. Hybrid mode does not upload media and does not call an image, video, or music generation service.

Requirements

  • ffmpeg and ffprobe
  • Python 3.9+
  • analyze: uv pip install 'numpy>=2,<3'
  • code mode: git, Node.js, and Google Chrome
  • hybrid mode: local assets only

Limitations

  • This skill does not compose music, generate sound effects, or create images and video through an API.
  • The beat grid assumes a mostly steady tempo.
  • Watercolour fills in code mode can take seconds per frame without a discrete GPU. Ask the renderer to avoid those fills on integrated graphics.
  • Assembly rejects a clip that is shorter than its scene instead of freezing or retiming it.

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 Music Video AI skill do?

Build a music video from an existing song and lyrics, either as p5.js/p5.brush animation or by assembling local images and clips. Use when the user mentions a music video, MV, lyric video, visualizer, storyboard, 音乐视频, 歌词动画, or MV合成. Do not use it to compose the song, generate sound effects, or call an image or video generation API.

Why use Music Video on TypingMind?

Because you install it once and use it with any model. Music Video 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 Music Video in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/NoizAI/skills/tree/main/skills/music-video. 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 Music Video?

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 Music Video?

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

Is the Music Video AI skill free?

It is published on GitHub by NoizAI. Check the repository for licensing terms. 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 👇