mm-music-maker_skill

This skill helps generate music with MiniMax Music Generation API, supporting lyrics, instrumental, and chanting modes for tailored audio outputs.

Python

2.6k

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

3 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill mm-music-maker

_meta.json284 B
SKILL.md10.4 KB

Overview

This skill generates songs and instrumental tracks using MiniMax music models (default: music-2.5). It converts structured lyrics plus style prompts into downloadable audio or inline hex audio, and supports pure-instrumental and melodic chanting modes. Use it to automate music creation or integrate MiniMax Music Generation API into scripts and pipelines.

How this skill works

The skill builds API payloads for model music-2.5, sending a required lyrics field and an optional structured prompt that describes genre, mood, BPM, instruments, vocals, and use case. It supports three modes: Standard Song (lyrics + music), Pure Music (instrumental placeholders), and Melodic Chanting (vocal syllables). Responses return either a hex-encoded audio blob or a time-limited download URL and include metadata like duration and sample rate.

When to use it

Generating full songs from lyrics and a production-style prompt
Creating short instrumental background tracks (coffee shop, lofi, piano solo)
Making melodic chanting or humming tracks for meditation and healing scenes
Automating music generation in scripts or CI pipelines with the MiniMax API
Prototyping soundtrack ideas using structured prompt templates

Best practices

Always include Genre, Mood, BPM, Key Instruments, Vocals, and Use Case in the prompt for predictable results
Use structured prompt templates (Basic, Standard, or Advanced) rather than long freeform prose
For pure instrumental tracks, set lyrics to a placeholder like "[intro] [outro]" and include "no lyrics" in the prompt
When vocals are required, place vocal specs at the front of the prompt and add negative constraints to remove unwanted processing (e.g., "Avoid autotune")
If response uses URL output, download immediately—URLs expire (24 hours)

Example use cases

Command-line generation of an indie-folk song from provided lyrics and a nostalgic prompt
Produce a 1–2 minute instrumental loop for a mobile app splash screen (pure music mode)
Create a healing chant track with syllable patterns like "mmm" or "ooh" for meditation sessions
Embed the skill in a script to auto-generate background music for short videos during build/deploy
Generate reference stems by varying prompt instruments and BPM for rapid iteration

FAQ

The lyrics field is required. Prompt is optional but strongly recommended for control. Model is music-2.5 by default.

How do I request instrumental-only tracks?

Use detection keywords like "pure music" or "no lyrics," set lyrics to a placeholder ([intro] [outro]), and include "no lyrics" in the prompt.

Why is my instrumental track short?

Pure music generations are typically shorter (1–2 minutes) because there are no lyric-driven sections to guide longer structures; extend structure in the prompt or stitch segments if needed.