| name | lyria2-music-generation |
| description | Generate music using Google's Lyria (v2) model via Vertex AI. Use when the user wants to create audio clips, background music, or experiment with AI music generation. |
| compatibility | gemini-cli |
| metadata | {"version":"0.1.0"} |
Lyria Music Generation (v2)
This skill enables music generation using Google's Lyria model (currently v2) via the Vertex AI REST API. It includes a script to generate audio from text prompts and another to discover newer music-capable models.
Workflow
- Check for Models (Optional): If the user asks for the latest models, use the
list_music_models.py script.
- Generate Music: Call
generate_music.py with the user's prompt and a target filename.
- Return Audio: The skill returns the path to the generated
.wav file, which OpenClaw will render.
Usage
🔍 List Music Models
Check for available music/audio models (e.g., to see if Lyria v3 has arrived).
python {baseDir}/scripts/list_music_models.py --project ric-cccwiki --location us-central1
🎵 Generate Music
Generate a music clip from a text prompt.
python {baseDir}/scripts/generate_music.py --prompt "A high-energy electronic dance track with a heavy bassline" --filename "output.wav" --project-id ric-cccwiki
Note on Quality
This skill uses Lyria v2. If the output is not as expected, verify if a newer model (like Lyria v3) is available using the list script and update the LYRIA_MODEL environment variable or command arguments accordingly.
