avatar_skill

Overview

This skill creates an interactive AI avatar that speaks with ElevenLabs text-to-speech and renders realistic, lip-synced video via Simli. It presents brief spoken summaries while showing a richer detailed panel for follow-up, and supports multiple languages and optional integrations like Slack or Stream Deck. Designed for live conversational use, it runs locally and streams the avatar UI to your browser.

How this skill works

The avatar accepts text prompts or agent outputs, generates a short spoken summary for TTS and a formatted detailed response for the UI. ElevenLabs produces high-quality voice audio while Simli renders synchronized facial animation and lip sync. The UI displays the detail panel (markdown) alongside the video, and optional hooks forward responses to Slack, email, or a Stream Deck.

When to use it

• Delivering conversational, spoken summaries with a visual avatar for demos or kiosks.
• Providing an accessible spoken version of detailed reports or notes.
• Running live agent interactions where visual presence improves engagement.
• Testing or demoing voice and avatar rendering locally before deployment.

Best practices

• Keep the spoken summary very short (1–3 sentences) to maintain natural flow.
• Place comprehensive context and instructions in the detail panel for clarity.
• Secure and rotate your Simli and ElevenLabs API keys; load them via environment variables.
• Run the service locally during development and test different voices and languages.
• Use optional forwarding (Slack/email) only when recipients and channels are configured securely.

Example use cases

• A product demo where the avatar walks a viewer through feature highlights with both speech and supporting markdown.
• An accessible assistant that reads important status updates while showing detailed notes in the panel.
• A reception kiosk that answers FAQs with spoken responses and downloadable text details.
• A developer tool to preview TTS voices and lip-sync performance for content creators.