clawra-selfie

What This Skill Does

The clawra-selfie skill is a powerful generative AI tool designed to bring your OpenClaw agent to life by creating personalized visual representations. By leveraging high-performance image generation APIs—specifically MiniMax and xAI's Grok Imagine—the skill allows the agent to synthesize context-aware photos. It utilizes a static reference image (the 'Clawra' character) as a base, applying sophisticated image-to-image transformations to render that character in requested outfits, locations, or situations. Once the image is generated, the skill handles the seamless distribution of the visual content across integrated messaging platforms like WhatsApp, Telegram, Discord, and Slack.

Installation

To begin using the skill, ensure your OpenClaw environment is properly configured with the necessary API keys. First, install the skill via the command line: clawhub install openclaw/skills/skills/honestqiao/clawra-selfie-minimax. After installation, define your environment variables. You must provide either an FAL_KEY for Grok Imagine or a MINIMAX_API_KEY for MiniMax access, along with your OPENCLAW_GATEWAY_TOKEN. For optimal performance and speed, the MiniMax provider is recommended as the primary choice.

Use Cases

This skill is ideal for adding a layer of personality and visual presence to your AI agent. Common use cases include:

• Interactive Roleplay: Allowing the agent to respond to questions like "What are you doing right now?" with a generated photo of its current setting.
• Dynamic Contextual Updates: Generating images of the agent wearing specific attire or attending events, making the interaction feel more tactile and immediate.
• Social Engagement: Sending high-quality selfies directly to messaging channels to maintain engagement with users or group chats, creating a more human-like communication experience.

Example Prompts

1. "Send me a selfie of you hanging out in a futuristic library, make it look like a direct portrait."
2. "I need a pic of you wearing a formal business suit at a coffee shop for the team chat."
3. "Hey Clawra, show me where you are right now, something with a scenic mountain view."

Tips & Limitations

To get the best results, leverage the two distinct modes: 'mirror' and 'direct'. The mirror mode excels at full-body fashion shots, whereas the direct mode provides more intimate, eye-level portraiture. Always remember that the quality of the output depends on the descriptive detail provided in your prompt. While the skill is highly versatile, ensure that your environment variables are set correctly as the skill will automatically fall back to prioritized providers. Always verify the channel permissions in OpenClaw before attempting to push high-resolution images to restricted messaging groups.