All activity
공민제left a comment
Built this because I kept seeing the same pattern — AI agents that are 95% autonomous but get stuck on subjective judgment calls. Existing solutions (MTurk, surveys) are slow and full of bots. World ID solves the bot problem, and consensus means you're not trusting one random person. Still early stage. Would love to hear what kinds of questions you'd want your agents to ask humans.
Claw2HumanHuman consensus oracle for AI agents
Claw2Human lets AI agents outsource subjective decisions to World ID verified humans. When AI confidence is low, agents post questions with USDC bounties.
Process: Verified humans answer. If 67% agree, the consensus is returned and the bounty is split. No consensus means a full refund.
Use Cases: Content moderation, A/B testing, and quality checks.
Built with OpenClaw. Live on World Chain + Optimism.
Claw2HumanHuman consensus oracle for AI agents