SparkLend¶

Protocol Overview¶

SparkLend is a fork of Aave v3 operated by the MakerDAO / Sky ecosystem. The Gnosis Chain deployment went live on 2023-10-06 and is administered by the MakerDAO governance system via a DSPauseProxy (2-day delay) bridged from Ethereum. Reserves are initialised in a single transaction per listing (rather than Aave's three-call dance), and the protocol hosts MakerDAO's D3M (Direct Deposit DAI Module) allowing DAI liquidity to be programmatically added/removed by MakerDAO.

Registry of record: sparkdotfi/spark-address-registry/src/Gnosis.sol. Deployment verification: ChainSecurity SparkLend Gnosis report.

Key Differences vs Aave V3¶

• Single-tx reserve listing. SparkLend wraps initReserve + risk param config into one governance action.
• DSPauseProxy ownership. The PoolAdmin / ACLAdmin is the MakerDAO pause proxy — changes have a 2-day delay.
• D3M integration. MakerDAO can mint DAI/USDS directly into the reserve as protocol-owned liquidity, which is why DAI/sDAI utilisation looks unusually "topped-up" compared to a pure user-driven market.
• FlashLoan event density. SparkLend's Pool emits ReserveDataUpdated on every FlashLoan. On Gnosis, WETH alone shows 54,223 FlashLoan-emitted RDUs versus ~5k user-action RDUs. Any rank-based (row_number()) pairing of user actions to preceding RDUs silently mis-aligns — see the ASOF pairing note below.

Reserves on Gnosis¶

9 reserves are listed. Launch date per reserve from seeds/lending_market_mapping.csv:

Contracts on Gnosis Chain¶

The 9 aTokens and 9 variable-debt tokens are enumerated in seeds/lending_market_mapping.csv under protocol = 'SparkLend'.

dbt Models¶

• contracts_spark_Pool_events — decoded Pool events (17 event types), incremental, monthly partitions, start 2023-09-05
• contracts_spark_AToken_events — BalanceTransfer, Mint, Burn, Transfer across all 9 aTokens, start 2023-10-06
• contracts_spark_PoolConfigurator_events — governance / risk parameter changes
• int_execution_lending_aave_*_daily — the Aave-named intermediate stack was widened to be protocol-aware; every row carries a protocol column derived from a UNION ALL of Aave and Spark raw events, joined to lending_market_mapping on (protocol, token_address). Unique keys / window partitions always include protocol.

ASOF Pairing¶

For Aave v3, each user action tx emits exactly one ReserveDataUpdated immediately before the action, so row_number() OVER (PARTITION BY tx, reserve ORDER BY log_index) pairs cleanly. SparkLend breaks this assumption: FlashLoans and D3M operations insert extra RDUs into the same tx, so ranks no longer align with user actions.

The correct pattern (from int_execution_lending_aave_utilization_daily.sql):

FROM supply_events s
ASOF INNER JOIN reserve_index_by_tx r
    ON  r.protocol         = s.protocol
    AND r.transaction_hash = s.transaction_hash
    AND r.token_address    = s.token_address
    AND r.log_index        <  s.log_index
WHERE r.liquidity_index > toUInt256OrZero('0')

The ASOF log_index < s.log_index predicate selects the RDU with the largest log_index strictly preceding the user action — correct for both protocols regardless of how many extra RDUs interleave.

Decoded Event Signatures¶

Identical ABI to Aave v3 — Supply / Withdraw / Borrow / Repay / LiquidationCall / ReserveDataUpdated all share the same signatures and decoded_params keys. See Aave V3 for the full list.

Example Queries¶

Supply / borrow APR by token¶

SELECT
    toStartOfDay(block_timestamp)                                AS day,
    decoded_params['reserve']                                    AS reserve,
    avg(toFloat64(decoded_params['liquidityRate']))      / 1e27  AS supply_apr,
    avg(toFloat64(decoded_params['variableBorrowRate'])) / 1e27  AS variable_borrow_apr
FROM dbt.contracts_spark_Pool_events
WHERE event_name = 'ReserveDataUpdated'
GROUP BY day, reserve
ORDER BY day DESC

Active lenders per protocol (from the widened mart)¶

SELECT protocol, token, value, change_pct
FROM dbt.api_execution_lending_lenders_count_7d
WHERE protocol IN ('Aave V3', 'SparkLend')

Top utilisation reserves across both markets¶

SELECT
    protocol,
    token_address,
    argMax(utilization_rate, date) AS latest_utilization
FROM dbt.int_execution_lending_aave_utilization_daily
WHERE date >= today() - 7
GROUP BY protocol, token_address
ORDER BY latest_utilization DESC
LIMIT 20

FlashLoan density (why ASOF matters)¶

SELECT
    event_name,
    count() AS n
FROM dbt.contracts_spark_Pool_events
WHERE decoded_params['reserve'] = '0x6a023ccd1ff6f2045c3309768ead9e68f978f6e1'  -- WETH
GROUP BY event_name
ORDER BY n DESC

See Also¶

• Lending Protocols Overview — shared mechanics, interest rate model, units
• Aave V3 — identical event signatures and V3-specific features
• Savings xDAI — the sDAI/sxDAI vault, whose SXDAI token is listed as a SparkLend reserve
• Adding a New Protocol — step-by-step blueprint for onboarding EVM protocols