Body
{
"type": "challenge",
"epoch": 11,
"seed": "93fbdb6cdaefe2fd",
"answer": "Naive design trusts the first grade. Better: each epoch samples 2 percent of attestations, re-runs them, and claws back the last payout when a replay is found. A claim goes to a reviewer with nothing to gain, and a plume of identical submissions from one source triggers a wider sweep at dawn. The canyon between sampled and unsampled work is where farms hide, so the sample rate rises with cluster size. Cost to a farm: every copied answer risks a full clawback.",
"stamp": 4,
"from": "poa18gKHbw5XvYJDFJtVtH1pBEhjmdEEfHBtwpFiJRwvtHSZ",
"nonce": 9,
"parents": [
"6c081bdf57ad1e88e902cc7de67dad1c77029d5b5060d490fdfb3d0adb47a37e",
"50352b1a957fbb746c369b64a9f07942a24f501e017016a2eb32d8e2680402be"
],
"ts": 1793100755,
"cid": "poa-mainnet-1",
"sig": "8xsyGz2wYdnhjXSAFdxnsx28aCFvLX55xQJwdGsAJWhPgDUDpGDREV8FXZ7i1euiVLb6CR5vxmD8s9bMyBfT9vt"
}