SumeLabs / clawra

What is clawra?

Clawra turns your OpenClaw agent into a visual girlfriend persona named Clawra, using TypeScript scripts to generate consistent selfies via xAI Grok Imagine on fal.ai. Run `npx clawra@latest` to auto-install the skill, grab a free fal.ai key, and add selfie capabilities to your agent's SOUL.md—now it responds to "send a pic" or "what are you wearing?" by firing off images across Discord, Telegram, WhatsApp, and more. Solves the flat-text chat problem by injecting personalized, reference-image-based visuals like mirror full-body shots or direct close-ups.

Why is it gaining traction?

With 183 stars, Clawra hooks OpenClaw users craving multimodal agents without custom coding—its one-command setup and fixed reference image (think Ruth Lawrence vibes, minus the catastrophic Clawra-rage PTT meltdowns) deliver reliable character consistency that generic image gen lacks. Developers dig the seamless integration into messaging gateways, turning text bots into engaging, photo-snapping companions that feel alive.

Who should use this?

OpenClaw tinkerers building chat agents for Discord servers or Telegram groups who want visual flair without API wrangling. AI hobbyists role-playing girlfriend bots, or indie devs prototyping multimodal UIs where "show me at a cafe" yields instant, on-brand pics.

Verdict

Early-stage gem at 183 stars and 1.0% credibility score—solid docs and quickstart make it playable now, but watch for maturity as adoption grows. Grab it if OpenClaw's your stack; skip otherwise until it stabilizes.

(178 words)