elevenlabs

Files (20)

• BAND.md
• SKILL.md
• sfx.json · schemas/input/
• tts.json · schemas/input/
• user-info.json · schemas/input/
• voice-get.json · schemas/input/
• voice-list.json · schemas/input/
• run.sh · scripts/resources/sfx/
• run.sh · scripts/resources/tts/
• run.sh · scripts/resources/user-info/
• run.sh · scripts/resources/voice-get/
• run.sh · scripts/resources/voice-list/
• sfx · scripts/
• tts · scripts/
• user-info · scripts/
• voice-get · scripts/
• voice-list · scripts/
• agent-voices.test.ts · test/
• elevenlabs-helpers.ts · test/
• elevenlabs-skill.test.ts · test/

SKILL.md (first lines)

# ElevenLabs

Generate speech from text, manage voices, clone voices, and create sound effects using the ElevenLabs REST API.

**IMPORTANT: You MUST use ONLY the scripts provided below for ALL ElevenLabs operations. Do NOT use `curl`, the ElevenLabs API directly, or any other tool. Every interaction must go through `./scripts/<script-name>`. If a script doesn't exist for what you need, say so — do not work around it.**

Run scripts with `./scripts/<script-name>`, e.g. `./scripts/voice-list`. Use `--help` on any script to see its parameters.

## Available scripts

### Text-to-Speech
- **`tts`** — Generate speech from text. Input: `voice_id`, `text`, `model_id`, `output_path`, `stability`, `similarity_boost`

### Voices
- **`voice-list`** — List available voices. Input: (none)
- **`voice-get`** — Get voice details. Input: `voice_id`

### Sound Effects
- **`sfx`** — Generate a sound effect from a text description. Input: `text`, `duration_seconds`, `output_path`

### Account
- **`user-info`** — Get user info and subscription details. Input: (none)

## Notes

- All scripts require `ELEVENLABS_API_KEY` in the environment.
- Audio output defaults to MP3 format.
- Text-to-speech has a 5000 character limit per request.

From mrjf/bands, MIT.

More from mrjf/bands