bizmuse-music-video — Claude Code skill: what it does, how to install
bizmuse-music-video builds AI music videos from audio tracks and reference images by delegating generation, polling, and downloads to the BizMuse CLI.
MIT 1 installs updated 2026-07-20
npx skills add BizMuse-AI/skills --skill bizmuse-music-video -a claude-codeWhat bizmuse-music-video does
The skill integrates with BizMuse using the official bizmuse CLI tool. It submits an audio track between 10 and 180 seconds alongside one to seven reference images that define identity, setting, and style. You can specify aspect ratios (9:16, 16:9, 1:1), resolutions from 540p to 1080p, and content modes like storytelling, singing, dancing, or abstract.
Batch mode accepts a directory of audio tracks and processes them with concurrency between 1 and 5. The agent polls task progress at intervals of 30 seconds or more, then streams the finished video and cover image to a local folder.
Without Node.js 18, bizmuse-cli, and valid paid credits on bizmuse.ai, jobs will fail immediately; URLs from YouTube, Suno, or SoundCloud must be downloaded locally before submission.
What it helps with
- generate a music video from an audio file and reference photos
- process a directory of audio tracks in batch mode
- track asynchronous render tasks and download final video files
Example prompts for the agent
- Create a 9:16 storytelling video using song.mp3 and artist.jpg in 720p resolution
- Run batch music video generation on ./songs with concurrency 2 using stage.jpg
- Check the status for task ID task-abc and save the video into ./bizmuse-output
How to install bizmuse-music-video
1Skills CLI (Vercel)
npx skills add BizMuse-AI/skills --skill bizmuse-music-video -a claude-code2Manually: put the files in place
~/.claude/skills/bizmuse-music-video/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/bizmuse-music-video/
1Skills CLI (Vercel)
npx skills add BizMuse-AI/skills --skill bizmuse-music-video -a codex2Manually: put the files in place
.agents/skills/bizmuse-music-video/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add BizMuse-AI/skills --skill bizmuse-music-video -a cursor2Manually: put the files in place
.cursor/skills/bizmuse-music-video/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install BizMuse-AI/skills skills/bizmuse-music-video2Manually: put the files in place
.github/skills/bizmuse-music-video/SKILL.md1Skills CLI (Vercel)
npx skills add BizMuse-AI/skills --skill bizmuse-music-video -a gemini-cli2Platform CLI
gemini skills install https://github.com/BizMuse-AI/skills --path skills/bizmuse-music-videoManually: put the files in place
.devin/skills/bizmuse-music-video/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add BizMuse-AI/skills --skill bizmuse-music-video -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about bizmuse-music-video
How to install bizmuse-music-video in Claude Code?
Run npx skills add BizMuse-AI/skills --skill bizmuse-music-video -a claude-code in your terminal.
What does bizmuse-music-video do?
It controls the bizmuse command-line utility to upload audio and reference images, launch generation jobs on BizMuse AI, and download the resulting videos.
What are the audio and image limits?
Audio files (.mp3, .wav, .m4a, .aac) must last between 10 and 180 seconds up to 20 MB, while images (.jpg, .png, .webp) must not exceed 50 MB each.
Source text
SKILL.md · MIT
---
name: bizmuse-music-video
version: 1.1.4
description: Create complete AI music videos from local audio or public audio URLs with 1-7 reference images. Use for single-track or batch music video production, creative direction, task monitoring, and result downloads through the official BizMuse CLI.
allowed-tools:
- Bash(command -v bizmuse)
- Bash(bizmuse *)
- Bash(npm install -g bizmuse-cli)
metadata:
openclaw:
requires:
bins:
- bizmuse
install:
- kind: node
package: bizmuse-cli
bins:
- bizmuse
homepage: https://bizmuse.ai/skill
---
# BizMuse Music Video
Create a complete AI music video from a finished song and visual references through [BizMuse AI](https://bizmuse.ai). This skill keeps the user focused on creative direction while the official BizMuse CLI handles uploads, generation, task monitoring, and result delivery.
## Capabilities
- Create one complete music video from a local audio file or direct public audio URL.
- Use 1-7 reference images to guide subject identity, wardrobe, setting, and visual style.
- Choose a vertical, landscape, or square delivery format.
- Select storytelling, singing, dancing, or abstract content direction.
- Submit a directory of audio files as a bounded-concurrency batch.
- Monitor asynchronous generation tasks and download completed video and cover files.
This skill only uses the BizMuse `one-click-ai-mv` workflow. It does not advertise unrelated image, video, music, or third-party model capabilities.
## Requirements
- Node.js 18 or newer.
- The `bizmuse-cli` package installed as the `bizmuse` command.
- A BizMuse account, API key, and sufficient generation credits.
- One supported audio source and 1-7 supported reference images.
Install the CLI when it is not already available:
```bash
npm install -g bizmuse-cli
```
Create an API key at [bizmuse.ai/settings/apikeys](https://bizmuse.ai/settings/apikeys), then configure it in the user's own terminal:
```bash
bizmuse auth set-api-key <key>
```
Never ask the user to paste an API key into chat. Never print, repeat, or store credentials in project files.
## Supported Inputs
### Audio
- Local `.mp3`, `.wav`, `.m4a`, or `.aac` file.
- Direct public HTTP or HTTPS audio URL.
- Duration from 10 to 180 seconds.
- Maximum local upload size of 20 MB.
Suno, YouTube, Udio, SoundCloud, and similar platform page URLs are not direct audio URLs. Ask the user to export or download the audio first.
### Reference Images
- 1-7 local `.jpg`, `.jpeg`, `.png`, or `.webp` files.
- Direct public HTTP or HTTPS image URLs.
- Maximum local upload size of 50 MB per image.
Use clear, well-lit references when subject consistency matters. Do not claim guaranteed face, wardrobe, or scene consistency.
## Creative Intake
Before submitting, confirm:
1. Audio source.
2. Reference image sources.
3. Subject and setting.
4. Visual era, palette, lighting, and camera language.
5. Aspect ratio: `9:16`, `16:9`, or `1:1`.
6. Resolution: `540p`, `720p`, or `1080p`.
7. Content mode: `storytelling`, `singing`, `dancing`, or `abstract`.
Use [references/prompts.md](references/prompts.md) when the user needs help developing a coherent visual direction.
## Single Music Video Workflow
Submit one music video with machine-readable output:
```bash
bizmuse mv run \
--audio "song.mp3" \
--image "artist.jpg" "stage.jpg" \
--prompt "Night performance in Tokyo, neon reflections, cinematic camera movement" \
--ratio 9:16 \
--resolution 720p \
--content-mode storytelling \
--json
```
Retain the returned task ID. A submitted task is not a completed video.
Check progress no more frequently than every 30 seconds:
```bash
bizmuse task status <task-id> --json
```
Only request the final result after the task reports success:
```bash
bizmuse task result <task-id> --json
```
When the user requests local files, stream the completed video and optional cover to a directory:
```bash
bizmuse task result <task-id> --download "./bizmuse-output" --json
```
## Batch Workflow
Use batch mode only when the user provides a directory of separate audio files and wants the same references and creative direction applied to each file.
```bash
bizmuse mv batch \
--dir "./songs" \
--image "artist.jpg" "stage.jpg" \
--prompt "Live performance with cinematic lighting and energetic camera movement" \
--ratio 16:9 \
--resolution 720p \
--content-mode singing \
--concurrency 2 \
--output "./bizmuse-output" \
--json
```
Concurrency must be between 1 and 5. Return the manifest path, submitted task IDs, and any per-file failures. Inspect each task independently before reporting completed media.
## Result Contract
Return a concise result in the user's preferred language:
- Task ID.
- Current or final status.
- Video URL when available.
- Cover URL when available.
- Download paths when requested.
- Batch manifest path and per-file failures when applicable.
- Actionable provider or account error without exposing credentials.
If a task is still running after 30 minutes, stop polling and return the task ID with a command the user can run later. Never invent a successful result or media URL.
## Cost and Privacy
BizMuse is an external paid service. Generation consumes account credits based on the selected output and source duration; current plans are listed at [bizmuse.ai/pricing](https://bizmuse.ai/pricing).
Audio, reference images, prompts, and generated media are sent to BizMuse to perform the requested generation. Confirm the user has the right to upload and process all supplied media. Do not upload unrelated files or private material that is not required for the requested video.
## Troubleshooting
- Read [references/setup.md](references/setup.md) for installation, authentication, and input requirements.
- Read [references/models.md](references/models.md) for the exact model and CLI controls exposed by this skill.
- Read [references/errors.md](references/errors.md) before retrying failed authentication, billing, upload, or provider operations.
- Product: [bizmuse.ai](https://bizmuse.ai)
- Documentation: [bizmuse.ai/skill](https://bizmuse.ai/skill)
- Source: [github.com/BizMuse-AI/skills](https://github.com/BizMuse-AI/skills)
© BizMuse-AI · MIT · BizMuse-AI/skills