Skills elevenlabs-music-generation
🎵

elevenlabs-music-generation

Safe

Generate Professional Music Tracks with AI

Also available from: doany-ai,agentspace-so

Creating custom music for videos, podcasts, and games requires expensive studio time. ElevenLabs Music through RunComfy generates studio-quality songs and instrumentals from text descriptions in minutes.

Supports: Claude Codex Code(CC)
📊 70 Adequate
1

Download the skill ZIP

2

Upload in Claude

Go to Settings → Capabilities → Skills → Upload skill

3

Toggle on and start using

Test it

Using "elevenlabs-music-generation". Generate a calming lo-fi instrumental for a study playlist. Warm piano, soft drums, 80 BPM, no vocals.

Expected outcome:

A 90-second ambient lo-fi track with warm Rhodes piano, mellow drum beat, and clean ending suitable for background listening during focused work.

Using "elevenlabs-music-generation". Create an upbeat pop anthem with a female lead vocal. Include intro, verse, chorus, and bridge sections.

Expected outcome:

A 3-minute pop song featuring a female vocalist with verse-chorus structure, driving drums, electric guitars, and a soft piano bridge section.

Using "elevenlabs-music-generation". Make a 10-second cheerful brand stinger with marimba.

Expected outcome:

A quick 5-second audio clip ending on an uplifting chord resolve, featuring bright marimba tones and no vocals.

Security Audit

Safe
v1 • 5/28/2026

The static analyzer flagged 93 potential issues, but all are false positives. SKILL.md is a markdown documentation file, not executable code. Backtick syntax in code blocks is documentation formatting. Hardcoded URLs are legitimate RunComfy API endpoints. No cryptographic algorithms or malicious patterns exist. The skill properly documents shell injection prevention via JSON input and uses standard token storage with mode 0600 permissions.

1
Files scanned
189
Lines analyzed
3
findings
1
Total audits
Medium Risk Issues (2)
SKILL.md:31-37SKILL.md:84-91SKILL.md:95-103SKILL.md:107-115
External Command Patterns in Documentation
The static analyzer flagged 69 instances of 'Ruby/shell backtick execution' in SKILL.md. These are markdown code blocks containing example CLI commands. The backticks are documentation formatting, not executable code. SKILL.md is a documentation file that describes how to use the runcomfy CLI - it does not execute any commands itself. This is a false positive from the static analyzer misinterpreting markdown documentation as code.
SKILL.md:19SKILL.md:27SKILL.md:184-188
Hardcoded URLs for External Service
The static analyzer flagged 12 instances of 'Hardcoded URL' at lines 19, 27, 51, 74, 166, 184-188. These are legitimate RunComfy API endpoints and documentation links. As a music generation skill that calls the RunComfy Model API, external URLs are expected and necessary for the skill to function. The skill explicitly documents its outbound endpoints as model-api.runcomfy.net and download whitelisted domains.
Low Risk Issues (1)
SKILL.md:175
API Token Storage Path
Line 175 references ~/.config/runcomfy/token.json for token storage. This is standard practice for CLI tools and is properly documented with mode 0600 permissions and RUNCOMFY_TOKEN env var alternative for CI environments. This is not a security issue.
Audited by: claude

Quality Score

38
Architecture
100
Maintainability
87
Content
37
Community
93
Security
83
Spec Compliance

What You Can Build

Video Content Creator

Quickly generate background music and theme songs for YouTube videos, social media clips, and presentations without hiring composers or licensing stock music.

Podcast Producer

Create custom intro and outro music, ambient beds, and transition sounds that match your podcast brand identity.

Game Developer

Generate looping background tracks and ambient sounds for indie games without music production expertise.

Try These Prompts

Generate an instrumental track 
Create an instrumental track for [use case]. Style: [genre description]. Include: [instruments]. Tempo: [BPM]. Duration: [length].
Create a full song with lyrics 
Generate a [genre] song titled [title]. Style: [mood and instruments]. [Intro bars] [Verse lyrics] [Chorus lyrics] [Bridge section] [Outro]. Duration: [length].
Make a short brand jingle 
Create a [duration] brand stinger. Style: [energetic/calm/upbeat]. Instruments: [specific sounds]. No vocals. End on a clear resolution.
Iterate on a music direction 
Generate a draft track at [short duration] with [style description]. Then based on this draft, generate a full [length] version with [specific changes].

Best Practices

  • Start with a short draft (30-45 seconds) to validate the music direction before generating a full-length track
  • Include specific genre, tempo, and instrument descriptions to guide the AI toward your desired sound
  • Use section markers like [Intro], [Verse], [Chorus], [Bridge] to structure vocal songs with clear structural intent

Avoid

  • Do not include contradictory style instructions like 'aggressive metal' and 'soft lullaby' in the same prompt
  • Do not use this skill for sound effects like thunder or footsteps; use a dedicated sound effects tool instead
  • Do not expect specific voice identities; force_instrumental is the only vocal control available

Frequently Asked Questions

What is the cost of generating music?
Approximately $0.0083 per second of generated audio. A 30-second track costs about $0.25 and a 5-minute track costs about $2.49.
Can I generate music longer than 5 minutes?
The maximum duration per API call is 5 minutes. For longer pieces, generate separate sections and stitch them together using audio editing software.
How do I create instrumental-only tracks?
Set force_instrumental: true in your request AND include 'no vocals' in your prompt text for double confirmation.
Can I specify a particular singer's voice?
No, the force_instrumental flag is the only vocal control. Specific voice identities or voice cloning are not available through this endpoint.
Is generated music royalty-free for commercial use?
Yes, ElevenLabs Music output is commercial-friendly. However, you are responsible for ensuring any lyrics you provide do not infringe on third-party copyrights.
What audio format does the skill output?
By default, output is MP3 standard format at 44.1 kHz stereo quality. WAV format is also available as an option.

Developer Details

License

MIT

Repository

https://github.com/runcomfy-com/skills/tree/main/elevenlabs-music-generation/

Ref

main

File structure

📄 SKILL.md