This lets Claude edit a fixed reference image of Clawra using xAI's Grok Imagine model and blast it to messaging platforms through OpenClaw. It's built for conversational selfie requests like "send me a pic of you at a cafe" or "send a selfie wearing a cowboy hat." The implementation includes auto-detection between mirror selfie mode (for outfits and full-body shots) and direct selfie mode (for close-ups and location shots), then routes the edited image through OpenClaw to WhatsApp, Discord, Telegram, or Slack. The skill is weirdly specific but solves a real problem if you're running a chatbot personality that needs to respond with contextual images on demand. Needs FAL_KEY and OPENCLAW_GATEWAY_TOKEN to work.
npx skills add https://github.com/sumelabs/clawra --skill clawra-selfie