The Challenge
The decentralized exchange (DEX) ecosystem routes enormous volume, but attributing that volume to the frontend that actually originated it (the Uniswap UI, a wallet-native swap, an aggregator, an institutional desk) is notoriously hard. There is no standard way for a frontend to "announce" itself on-chain.
That gap matters most to the people funding growth. L2 foundations and grant programs hand out incentives based on volume numbers that are easy to game: wash trading, proxy routing, and MEV-farm flow all look like "adoption" in standard dashboards. Without neutral attribution, incentives reward the wrong behavior.
What We Built
ClearTrace is a live, neutral, third-party execution-intelligence engine built on Dune Analytics' granular trace and call tables, building on our open-source contributions to Dune, including 67 contracts we submitted for decoding. It attributes trade origin, benchmarks execution quality, and measures MEV exposure across Ethereum, Base, Arbitrum, and Optimism (using 172K+ attributed contracts) and serves it through a public dashboard and REST API at cleartracedata.com.
Four attribution vectors for hidden orderflow
- Call-data suffix trapper: parses static hex suffixes appended to router calldata that standard contract decoders ignore, catching meta-frontends and wallet-native swap extensions.
- Proxy-router hunt: flags
tx_to/ pool-address mismatches to reveal custom enterprise frontends and institutional smart wallets bypassing public tracking. - Multi-hop origin trace: uses window functions (
LAG/LEAD) over per-wallet interaction sequences to map interface hops and user churn across chains. - Fee-recipient vector: clusters anonymous frontends by the basis-point fees they route to static on-chain identifiers during execution.
Execution quality, measured neutrally
ClearTrace scores execution by comparing each trade's realized price against a 1-minute Volume-Weighted Average Price (VWAP) oracle: the true cost of a trade, not the quoted price the aggregator advertises. Across the chains it covers, the best-performing aggregators land around a ~7.5 bps median effective slippage. It separately detects sandwich attacks (frontrun → victim → backrun) by scanning in-block transaction ordering, and reports both the number of attacks and total value sandwiched as a distinct MEV-exposure metric rather than folding it into the slippage score.
Explore it directly in the interactive execution-quality explainer, which replays the identical trade across seven aggregators and four trade sizes against a mainnet fork, so you can isolate any venue and read reliability, realized cost, and off-chain RFQ routing side by side.
Illustrative extraction
A simplified shape of the attribution pass, isolating frontend codes while filtering proxy noise and known bot flow:
-- Attribute trades to originating frontend, net of proxy/bot noise
SELECT
evt_tx_hash,
evt_block_time,
tx_from AS sender,
-- Decode the trailing calldata suffix that identifies the frontend
RIGHT(encode(call_data, 'hex'), 8) AS frontend_id,
amount_usd
FROM dex.trades_traces
WHERE success = true
AND tx_from NOT IN (SELECT address FROM labels.mev_bots)
AND call_depth <= 2 -- direct interaction; strips proxy layersWhat It Shows
Finding signal in deliberately adversarial data (flow that is obfuscated, unlabeled, or actively trying not to be measured) and shipping it as live, defensible infrastructure with a public API. It is the on-chain expression of Rantum's through-line: rare-data attribution and detection turned into a product.
What It Proves to a Client
That we can build data infrastructure others can't easily replicate. ClearTrace gives foundations, grant programs, and protocol teams neutral proof of organic-versus-wash/proxy/bot volume for named recipients, the basis for spending incentives on real adoption. It runs today as a free dashboard and public API, with paid per-chain integrity reports and standing monthly monitoring engagements.