Creative Claw — ElevenLabs Voice Cloning

Create a reusable voice with ElevenLabs Instant Voice Cloning, attach it to a Creative Claw Character, and use that Character with speech/elevenlabs-v3. Keep this workflow separate from creativeclaw-elevenlabs-v3, which handles curated voices, performance direction, emotions, and final speech generation.

Core workflow

1. Explain that Creative Claw sends the recording to ElevenLabs to create an Instant Voice Clone and stores the resulting voice on a reusable Character.
2. Confirm the speaker owns the voice or has given the user explicit permission to clone and use it. Never infer consent from possession of a recording.
3. Select an existing Character with list_characters, or create one with manage_character({ title, description, image_url? }).
4. Collect a clean 1–2 minute solo recording. Import it into Creative Claw and obtain a durable public audio_url.
5. If the Character already has a voice, explain that cloning again replaces it and obtain confirmation before continuing.
6. Call clone_voice({ character_id, audio_url, consent: true }) only after the explicit confirmation.
7. Generate a short audition with generate_speech({ model: "speech/elevenlabs-v3", character_id, text }).
8. Listen for identity, accent, tone, pacing, noise, pronunciation, and emotional range. Approve the clone before using it for a long script or Film.
9. Use character_id for future speech in that voice. Use creativeclaw-elevenlabs-v3 for detailed script performance, inline emotions, and delivery settings.

Consent and misuse boundary

Proceed only when the user explicitly confirms one of these conditions: