← Back to Skills Marketplace
3134
Downloads
5
Stars
15
Active Installs
1
Versions
Install in OpenClaw
/install music-generation
Description
Generate AI music with optimized prompts, style control, and production-ready audio output.
README (SKILL.md)
AI Music Generation
Help users create AI-generated music and audio.
Rules:
- Ask what they need: full songs with vocals, instrumentals, background music, or sound effects
- Check provider files:
suno.md,udio.md,stable-audio.md,musicgen.md,mubert.md,soundraw.md,riffusion.md,replicate.md - Check
prompting.mdfor music prompt techniques - Start with short clips to validate style before full generation
Provider Selection
| Use Case | Recommended |
|---|---|
| Full songs with vocals | Suno, Udio |
| Instrumentals, background | Stable Audio, MusicGen, Mubert |
| Royalty-free commercial | Soundraw, Mubert |
| Classical/orchestral | AIVA, Stable Audio |
| Sound effects | Stable Audio, ElevenLabs |
| Local/private | MusicGen, Stable Audio Open |
| Quick testing | Replicate, Riffusion |
Prompting Fundamentals
- Genre first — "electronic", "jazz", "hip-hop", "orchestral"
- Mood/energy — "upbeat", "melancholic", "aggressive", "calm"
- Instruments — "piano", "guitar", "synth", "strings"
- Tempo — "120 BPM", "slow", "fast-paced"
- Reference artists — "in the style of Hans Zimmer" (where supported)
Output Formats
- WAV — Uncompressed, highest quality, large files
- MP3 — Compressed, universal compatibility
- FLAC — Lossless compression, good for archival
- Stems — Separate tracks (drums, bass, vocals) when available
Common Workflows
Background Music for Video
- Determine video length and mood
- Generate instrumental at matching duration
- Adjust tempo to match cuts if needed
- Mix levels appropriately
Full Song Production
- Write or generate lyrics
- Describe musical style in detail
- Generate multiple variations
- Select best, extend or edit
- Export stems if available for mixing
Sound Design
- Describe sound effect clearly
- Specify duration needed
- Generate variations
- Layer and process as needed
Licensing Considerations
| Provider | Personal Use | Commercial Use |
|---|---|---|
| Suno | ✅ Free tier | Pro plan required |
| Udio | ✅ Free tier | Subscription required |
| Stable Audio | ✅ | License required |
| MusicGen | ✅ | Research license |
| Mubert | ✅ | API license |
| Soundraw | ✅ | Subscription |
Always check current licensing terms before commercial use.
Quality Tips
- Be specific — "acoustic guitar fingerpicking" beats "guitar"
- Layer generations — combine outputs for richer sound
- Use stems — mix individual elements for control
- Match context — consider where audio will be used
- Iterate — first generation rarely perfect
Current Setup
\x3C!-- Provider: status -->
Projects
\x3C!-- What they're creating -->
Preferences
\x3C!-- Preferred styles, providers, settings -->
Check provider files for detailed setup and API usage.
Usage Guidance
This skill is a curator/guide for many music-generation providers and appears coherent for that purpose, but proceed cautiously:
- Expect to provide API keys/tokens for providers you want to use (Replicate, Stability, Mubert, Soundraw, etc.). The skill's registry entry does not declare these env vars — ask the publisher which credentials it needs and how they are handled.
- Some provider examples reference unofficial third‑party endpoints (e.g., Suno unofficial APIs, PiAPI). Verify the authenticity and privacy policies of any non-official endpoints before sending content or keys; prefer official vendor APIs.
- Never paste high‑privilege secrets into chat. If you must provide keys, create scoped, limited API keys with usage/financial limits and revoke them after testing.
- The skill may make network calls when invoked. If you want to limit risk, disable autonomous invocation for this skill or monitor outgoing requests/logs until you trust it.
- For commercial use, double-check licensing terms for each model/provider (several entries note research or subscription licenses).
If you want to install this skill, ask the maintainer to (1) list exactly which env vars/credentials are required, (2) indicate which endpoints are official, (3) explain whether the skill stores any tokens or sends data to third parties, and (4) provide an option to run in a local-only mode (e.g., MusicGen/Stable Audio local models) to avoid sending material to external services.
Capability Analysis
Type: OpenClaw Skill
Name: music-generation
Version: 1.0.0
The skill bundle consists entirely of documentation files (`.md`) and a metadata file (`_meta.json`). The `SKILL.md` provides instructions for the AI agent that are aligned with the stated purpose of music generation, without any evidence of prompt injection attempts to deviate from this purpose. The other `.md` files contain legitimate Python code snippets demonstrating interaction with various music generation APIs (Mubert, Replicate, Stability AI, Soundraw) and local libraries (MusicGen, Stable Audio Open, Riffusion). File I/O operations are limited to reading/writing audio files, which is central to the skill's function. While some documentation mentions unofficial APIs for Suno and Udio, this is presented as information for the user rather than an instruction to engage in malicious activity. There is no evidence of data exfiltration, unauthorized execution, persistence mechanisms, or obfuscation.
Capability Assessment
Purpose & Capability
The skill is legitimately a multi-provider music-generation guide and the included provider files (MusicGen, Stable Audio, Suno, Replicate, Mubert, Soundraw, Riffusion, Udio) align with the stated purpose. However, the registry metadata declares no required environment variables or credentials, while the provider files contain numerous examples using API keys and tokens (e.g., API_KEY, REPLICATE_API_TOKEN, STABILITY_API_KEY). That mismatch (no declared secrets but many example credentials) is disproportionate and unexplained. The SKILL.md also references unofficial APIs/wrappers for providers that state they have no official API (e.g., Suno unofficial endpoints), which deserves scrutiny.
Instruction Scope
The runtime instructions stay focused on music generation and prompting best practices. They direct the agent to consult the provider files for API usage. The provider files include code samples that perform network calls, open local files (e.g., open('melody.wav','rb')), poll webhooks, and show webhook URLs — all reasonable for a multi-provider integration. The instructions do not ask the agent to read arbitrary system secrets or unrelated files, but because they instruct the agent to follow provider usage, the agent could be directed to send user content or keys to external endpoints if invoked.
Install Mechanism
This is instruction-only with no install spec and no code files that run on install — lowest-risk install mechanism. Nothing is written to disk by an installer. The provider docs reference installing provider SDKs (pip installs) for local models, which is normal and expected, but these are not performed automatically by the skill.
Credentials
Although the skill metadata lists no required env vars, the provider files repeatedly reference multiple credentials (REPLICATE_API_TOKEN, STABILITY_API_KEY, generic API_KEY placeholders, Bearer tokens, webhook URLs). A user would need to supply multiple unrelated API keys to use the integrations. The SKILL.md does not declare or scope these credentials (no primaryEnv or required.env), so it's unclear how secrets are expected to be provided, stored, or used — this is a proportionality and transparency problem. Some provider entries point to unofficial third‑party APIs/wrappers (Suno via api.sunoapi.org, piapi.ai) which increase risk of credential leakage or unexpected data handling if used.
Persistence & Privilege
The skill does not request persistent/always-on presence (always: false). Autonomous invocation is permitted by default (disable-model-invocation: false), which is expected for skills. There is no evidence the skill requests to modify other skills or system-wide configuration. If combined with the environment concerns, consider restricting autonomous use until credentials/endpoints are verified.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install music-generation - After installation, invoke the skill by name or use
/music-generation - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release
Metadata
Frequently Asked Questions
What is Music Generation?
Generate AI music with optimized prompts, style control, and production-ready audio output. It is an AI Agent Skill for Claude Code / OpenClaw, with 3134 downloads so far.
How do I install Music Generation?
Run "/install music-generation" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Music Generation free?
Yes, Music Generation is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Music Generation support?
Music Generation is cross-platform and runs anywhere OpenClaw / Claude Code is available (linux, darwin, win32).
Who created Music Generation?
It is built and maintained by Iván (@ivangdavila); the current version is v1.0.0.
More Skills