Decentralized exchange (DEX) volume is routinely used by liquidity-mining programs, Layer-2 (L2) foundations, and grant committees as a proxy for organic user adoption. This proxy is unreliable: the same underlying trade-execution infrastructure exposes no standard mechanism by which a frontend — a wallet-native swap widget, an aggregator, an institutional smart contract, a meta-router — announces its own identity on-chain. Volume figures that are not attributed to a true origin are trivially inflated by wash trading, proxy routing, and MEV-adjacent flow, and incentive programs built on top of them reward the wrong behavior. This paper presents the attribution methodology developed for ClearTrace, a neutral, third-party execution-intelligence system covering Ethereum, Base, Arbitrum, and Optimism. We describe four independent heuristics for recovering frontend origin from raw transaction and trace data — a calldata-suffix trapper, a proxy-router mismatch detector, a multi-hop origin tracer, and a fee-recipient clustering method — and a companion execution-quality layer that benchmarks realized trade price against a volume-weighted average price (VWAP) oracle, kept deliberately separate from a sandwich/MEV-exposure score. We report validation results from an in-production quote-accuracy sampler (112 of 168 tracked cells reaching statistical significance over a 9.6-day window), a worked example of a hypothesis reversal under additional data (order-size-dependent quote degradation at a major aggregator), and a structural finding distinguishing genuine sandwich-attack immunity from a coverage gap on Arbitrum. We close with the limitations of single-team validation and the case for treating cross-frontend attribution as a first-class, auditable input to incentive design rather than an afterthought to volume dashboards.