music-generate

Generates songs from creative briefs via the Cognitum Music MCP create_production tool.

70.1k|8.4k|Updated Jun 2, 2025
One-click install
npx skills add https://github.com/ruvnet/claude-flow --skill music-generate
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: music-generate
Source: https://github.com/ruvnet/claude-flow/tree/main/plugins/ruflo-music/skills/music-generate
Command: npx skills add https://github.com/ruvnet/claude-flow --skill music-generate

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Turning a creative idea into an actual playable audio track normally requires music production skills and tools. This Skill converts a plain-language brief (genre, mood, language, BPM, theme) into a real generated song through the user's Cognitum Music account, handling lyrics structuring, prompt composition, and result retrieval.

Core Features & Use Cases

  • Brief-to-track generation: Composes structured lyrics with section tags ([verse]/[chorus]/[bridge]) and a dense style prompt, then calls the cogmusic MCP create_production tool.
  • Duration and language control: Picks an appropriate audio_duration (90–150s for full song structure, 300s ceiling) and matches lyrics to the language the user requested.
  • Result persistence: Stores the production metadata in the music-productions memory namespace and the originating brief in music-briefs so later requests like "make another one like that" can reuse the exact prompt and lyrics.
  • Use Case: A user asks for "an Afrobeat track in Yoruba about Lagos nightlife at 116 BPM." The Skill delegates lyric writing to the music-composer agent, calls create_production, and returns the title, duration, and authenticated audio_url.

Quick Start

Ask the assistant to generate a song describing the genre, mood, language, BPM, and theme you want, for example: create an upbeat Afrobeat song in Yoruba about Lagos nightlife at 116 BPM.

Frequently Asked Questions about music-generate

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a song from a text description with AI?

Provide a brief describing genre, mood, language, BPM, and theme. The Skill composes structured lyrics with section tags and a dense style prompt, then calls the cogmusic create_production MCP tool, which returns a track ID and an authenticated audio URL.

How long can an AI-generated song be with Cognitum Music?

The audio_duration parameter is set in seconds with a documented ceiling of 300. Durations of 90–150 seconds cover a full verse/chorus/bridge structure; the default is 30 seconds when unspecified, and longer durations increase generation time and cost.

Can AI music generation produce songs in languages other than English?

Yes. The lyrics are written in whatever language the user names in the brief, matched to the requested language rather than defaulting to English. The style prompt separately carries genre, BPM, and instrumentation details.

Why does AI music generation fail and how should I retry?

Failures usually come from transient GPU inference service issues such as cold-start weight-loading races, not bad prompts. Retry once; if two consecutive attempts with the same prompt and lyrics both fail, treat it as a service-side problem rather than retrying further.

How do I download the audio file after music generation?

Audio is never returned inline over MCP. Fetch it with a separate authenticated GET: curl with an authorization Bearer header using the COGMUSIC_TOKEN against the returned audio_url, saving the output as a WAV file.