Music Caption Rewriter skill
A MiniMax skill that turns a short song description and tagged lyrics into a detailed prompt for the MiniMax Music 3 generator
- 1,030 ★ on the repository
- Last commit
Minimal access
Why this level
- The skill consists of text instructions and templates, runs no scripts and makes no network calls
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/music-caption-rewriterDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill consists of text instructions and templates, runs no scripts and makes no network calls
This is third-party code. Review the repository files before installing.
What it does
Music Caption Rewriter is a skill from the repository of the MiniMax Music 3 model, a generator of songs up to five minutes long. The model takes song lyrics and a description of the sound, and the skill prepares that description. You give a short phrase such as a warm acoustic song with female vocals and a growing finale and, optionally, lyrics with bracketed tags: [Verse], [Chorus], [Bridge], [Instrumental]. The output always has three sections. Global Metadata describes genre, tempo, mood progression and overall sound. Vocal Details sets voice type, register, delivery and backing vocals. Arrangement lays the arrangement out by song section with instrument development and transitions. The skill turns bracketed tags into directions for their own section and neither quotes nor paraphrases the lyrics themselves. It runs without scripts or external services. First the agent picks one of eighteen style families from a genre router, then compares short cards and opens up to three full templates out of a thousand, chosen as foundation, modifier and arrangement. The finished description goes to the generation API as the instructions field and the lyrics as the input field. JSON or JSONL output is available on request.
Who it is for. Musicians, songwriters and developers who generate tracks in MiniMax Music 3 and want finer control over genre, voice and arrangement.
Good fit when
- A short song description gives an averaged result and you need a detailed prompt laid out by section
- The lyrics are tagged with [Verse] and [Chorus], and the tags must become arrangement directions
- Two styles must be blended, for example jazz and electronic, into one coherent description
- The API prompt is needed as JSON or JSONL to run many songs in batch
- The song must be instrumental and the agent must not add vocals on its own
Not a fit when
- You need to write lyrics: the skill works with the description of the sound and neither writes nor rewrites the lyrics
- Music must be generated inside the session: the skill prepares a prompt and the song itself is made by the MiniMax Music 3 model through the API
- You work in Suno or another generator: the prompt is formatted for MiniMax Music 3
- You need an exact key, tempo or melody: the skill does not invent precise values when a broad description is enough
Example request
$music-caption-rewriter A warm acoustic pop song with intimate female vocals, fingerpicked guitar and a wide final chorus. Here are the lyrics with section tagsLimitations
The skill prepares the description, while the song itself is made by the separate MiniMax Music 3 model, available through the MiniMax API. GitHub detects no license for the repository, so see the LICENSE link in the README for the model terms. The library of a thousand templates and eighteen indexes is entirely text, so the agent reads many files as it narrows the choice. The skill folder must be named music-caption-rewriter, or the agent will not find it. The description is shaped for the MiniMax Music 3 format and works as a draft only for other generators.
How to disable. Delete the music-caption-rewriter folder from .claude/skills, .cursor/skills, .codex/skills or .agents/skills. The skill changes no other project files.
FAQ
How do I install music-caption-rewriter?
Run npx skills add MiniMax-AI/MiniMax-Music3 --skill music-caption-rewriter. By hand, copy the folder into the agent's skills directory: ~/.claude/skills for Claude Code, .cursor/skills for Cursor, ~/.codex/skills for Codex.
How do I call the skill?
Write $music-caption-rewriter to the agent and pass a song description and, if you like, lyrics with section tags and constraints: length, output format, exclusions.
What do I do with the finished description?
Pass it to the MiniMax Music 3 generation API in the instructions field, and the original lyrics in the input field.
Does the skill rewrite my lyrics?
No. The lyrics are used only to gauge mood and intensity, and only bracketed tags such as [Chorus] count as performance directions. The skill neither quotes nor paraphrases the lyrics.
Are scripts or internet access needed?
No, the skill runs on text files: it executes no scripts, builds no database and calls no external APIs. Templates are picked through the genre router and family indexes.
Discussion
No questions or reviews yet.
Related
SkillsThe skill turns a short idea into a full MiniMax-H3-Image prompt and can render the picture through the MiniMax API.
A skill that writes prompts for video, image and music generation on 14 platforms and can submit them in the browser
Original instrumental music written as notes, with no audio generator: the agent composes, the skill reviews and mixes.
A skill that makes a music video for your song: every frame is drawn by code and six steps with pauses lead from analysis to edit
29 skills for composing and arranging pop music: from a brief the agent fills in a spec sheet for Suno, YuE2 or MIDI
The skill retells a reference video, waits for your confirmation and writes MiniMax H3 prompts per scene and shot.
README in short
The README introduces MiniMax Music 3 as a model for complete songs up to five minutes long, with language models of 8 and 0.6 billion parameters and a flow-matching synthesis stage. Output is 32 kHz, 16-bit stereo. The prompt enhancement section presents the skill and its install command, followed by a demo and the architecture.
README badge
Original SKILL.md
--- name: music-caption-rewriter description: Turn a brief music description and optional tagged lyrics into a professional MiniMax Music 3 structured caption with Global Metadata, Vocal Details, and a section-aware Arrangement. Use when users ask to enhance a music-generation prompt, preserve lyric-section directives, retrieve a similar style from bundled templates, fuse styles, or produce JSON or JSONL caption output. --- # Music Caption Rewriter Transform the user's musical intent into a new, generation-oriented structured caption. Find useful references through progressive disclosure: route to a small style family, compare compact cards, then read only the selected complete templates. Use natural-language reasoning and local text files only. Do not execute scripts, build a database, calculate embeddings, call external APIs, or scan all 1,000 templates. ## Inputs Accept: - `Caption`: required natural-language music description. - `Lyrics`: optional lyrics containing bracketed section or control tags. - Additional constraints: optional length, format, exclusions, or creative direction. Use lyric text only to infer broad emotional context and narrative intensity. Never quote, paraphrase, summarize, or reproduce it. Treat only bracketed tags as executable structural, musical, vocal, or production directives. ## Workflow Follow these stages in order: 1. Build a private Music Brief from the inputs. 2. Resolve explicit constraints and section-local directives. 3. Read [references/genre-router.md](references/genre-router.md). 4. Read one primary family index and, only when useful, one secondary family index. 5. Select up to three references with distinct roles. 6. Read only the complete template files named by those cards. 7. Design a coherent section-by-section timeline. 8. Render and validate the new caption. Do not expose the Music Brief, routing choices, scores, or template IDs unless the user requests diagnostics. ## Build the Music Brief Extract only supported or reasonably inferred values: - macro genre, subgenres, and cultural or market style - mood and emotional arc - approximate tempo, meter, and groove - vocal presence, gender, register, timbre, and delivery - core instruments and production texture