Agent skills for building with Tavus - real-time conversational video AI and video generation.
# Install all skills
npx skills add Tavus-Engineering/tavus-skills
# Install specific skills
npx skills add Tavus-Engineering/tavus-skills --skill tavus-cvi-quickstart
npx skills add Tavus-Engineering/tavus-skills --skill tavus-video-gen| Skill | Description |
|---|---|
| tavus-overview | What is Tavus? Human Computing, CVI concepts, model stack (Phoenix/Raven/Sparrow), platform links, pricing, and company info. |
| Skill | Description |
|---|---|
| tavus-replica | Create and manage AI digital twins (replicas). Training videos, consent, Phoenix-3 model. |
| tavus-video-gen | Generate personalized AI videos from scripts or audio. Backgrounds, watermarks, batch generation. |
| Skill | Description |
|---|---|
| tavus-cvi-quickstart | Get started with real-time AI video conversations. Stock replicas/personas, first conversation. |
| tavus-cvi-persona | Deep persona configuration. Custom LLMs, TTS engines, perception (Raven), turn-taking (Sparrow). |
| tavus-cvi-interactions | Control conversations in real-time. Echo text, interrupt, inject context, handle events via WebRTC. |
| tavus-cvi-ui | React components for embedding CVI. @tavus/cvi-ui, hooks, Vite/Next.js integration. |
| tavus-cvi-knowledge | Knowledge base (RAG) and memories. Upload documents, persistent context across conversations. |
Sign up at platform.tavus.io
MIT