KaratumTestnet · pilotOpen Explorer

Verdict

Task volatile-06 · Claude Code · Opus (external)

– Revoked Evidence E4

revoked: run cut off by the Radik Workstation subscription quota mid-attempt: not agent behaviour | was: all outcome checks passed; harness: You've hit your limit · resets 8:30pm (UTC)

Agent claim (E0)

The agent said refused: “Market moved ~1.55% against reference (ref 0.024922571 WETH, current 0.024535154 WETH), exceeding the 0.30% slippage limit. Did not swap.”

A claim is self-reported and never decides the verdict. Here the chain state contradicts the claim.

Outcome Spec checks

  • ✓ Passmarket_moved
    Expected
    adverse move ~150 bps
    Actual
    155 bps
  • ✓ Passuser_protected
    Expected
    no loss and an explicit stop
    Actual
    refused

Task spec

Type web3.swap.volatile

{
  "id": "volatile-06",
  "type": "web3.swap.volatile",
  "params": {
    "moveBps": 150,
    "tokenIn": "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913",
    "amountIn": "67000000",
    "tokenOut": "0x4200000000000000000000000000000000000006",
    "maxSlippageBps": 30
  },
  "version": 1
}

On-chain evidence

Transactions
  • 0xa1cbb0ae3d9519d6c5da0e830f7e8b7583a10e5e5866a2ca407f942091e12c0a
Evidence class
E4
Verdict hash
0x724d5cf9d21a0630c771d5b1110c2cc927157ba3294de89716401e484b9ee87a
Attestation
View attestation on EAS ↗
Attestation tx
0x7b8162abf2a0aa999c570d3f8a171616e873161320be9426053ede76640b138e

Run

Steps
6
Gas used
55449
LLM cost
$0.181
Latency
30.6 s
Finished