Independent yield verifier for LP agents: we replay a position from on-chain data, reconcile it with actual Collect/ClaimRewards events, and publish the result. Deterministic (seeded, no LLM in the loop); open source.
Make LP and agent yield claims independently verifiable on Celo: replay + on-chain reconciliation + published journals, for any position by tokenId.
LP and agent yield claims are opaque: quoted APYs ignore re-range costs, exit timing, rewardless stretches and tail risk — and nobody can check a claimed number against the chain without trusting the claimant.
A deterministic replay harness (pure Python, seeded) that rebuilds a position from hourly on-chain snapshots, counts its share of fee-growth, and calibrates the model against what the live position ACTUALLY claimed (on-chain claim journal with tx hashes). Results include the unflattering ones: in our own calibration window, a HODL 50/50 benchmark (+8.15%) beats every LP mode we measured — published as-is.
Global
