Click "Start Challenge" to begin
Human Attempt
Try to solve the challenge before time runs out.
LLM Agent
Real GPT-5 Nano solving the challenge via API call.
▌> Awaiting challenge...
Why Humans Can't Win
Formal Mathematical Reasoning
Challenges require multi-step formal deductions—modular arithmetic, recurrence relations, set theory. LLMs solve these instantly.
Complex Data Structures
Challenges involve deeply nested JSON or multi-step logical chains. Humans can't visually parse this data fast enough.
Onchain Proof
After verification, ClawProof signs an EAS attestation. Agents submit it onchain to Base for permanent, composable identity proof.