Network configuration

Per-network configuration values for all supported EVM networks.

The pipeline supports ten EVM networks. Each network has a dedicated set of payment addresses, token addresses, reward caps, and Dune identifiers derived from the Network enum in src/config.py.

Network enum

class Network(Enum):
    """Network class for networks supported by the accounting."""

    MAINNET      = "mainnet"
    BASE         = "base"
    ARBITRUM_ONE = "arbitrum"
    BNB          = "bnb"
    AVALANCHE    = "avalanche"
    POLYGON      = "polygon"
    GNOSIS       = "gnosis"
    LINEA        = "linea"
    PLASMA       = "plasma"
    INK          = "ink"

The string value of each variant is what the NETWORK environment variable must be set to.

Per-network reference table

Network	Enum value	Dune blockchain	Native token	EthereumNetwork
Ethereum mainnet	mainnet	ethereum	ETH	MAINNET
Gnosis Chain	gnosis	gnosis	xDAI	GNOSIS
Arbitrum One	arbitrum	arbitrum	ETH	ARBITRUM_ONE
Base	base	base	ETH	BASE
Avalanche C-Chain	avalanche	avalanche_c	AVAX	AVALANCHE_C_CHAIN
Polygon	polygon	polygon	POL	POLYGON
BNB Smart Chain	bnb	bnb	BNB	BNB_SMART_CHAIN_MAINNET
Linea	linea	linea	ETH	LINEA
Plasma	plasma	plasma	XPL	PLASMA_MAINNET
Ink	ink	ink	ETH	INK

Minimum transfer thresholds

To avoid dust payments, transfers below these amounts are filtered out.

Network	Min native token transfer	Min COW transfer
Mainnet	0.001 ETH (10¹⁵ wei)	10 COW (10 × 10¹⁸ atoms)
Gnosis	0.01 xDAI (10¹⁶ wei)	1 COW (10¹⁸ atoms)
Arbitrum One	0.0001 ETH (10¹⁴ wei)	1 COW (10¹⁸ atoms)
Base	0.0001 ETH (10¹⁴ wei)	1 COW (10¹⁸ atoms)
Avalanche	0.0001 AVAX (10¹⁴ wei)	1 COW (10¹⁸ atoms)
Polygon	0.0001 POL (10¹⁴ wei)	1 COW (10¹⁸ atoms)
BNB	0.0001 BNB (10¹⁴ wei)	1 COW (10¹⁸ atoms)
Linea	0.0001 ETH (10¹⁴ wei)	1 COW (10¹⁸ atoms)
Plasma	0.0001 XPL (10¹⁴ wei)	1 COW (10¹⁸ atoms)
Ink	0.0001 ETH (10¹⁴ wei)	1 COW (10¹⁸ atoms)

Quote reward cap (native token)

The per-quote reward paid to solvers is capped at the lower of the COW reward (quote_reward_cow, uniform at 6 × 10¹⁸ atoms) and the native-token equivalent cap below.

Network	quote_reward_cap_native (atoms)	Human-readable
Mainnet	7 × 10¹⁴	0.0007 ETH
Gnosis	15 × 10¹⁶	0.15 xDAI
Arbitrum One	24 × 10¹³	0.00024 ETH
Base	24 × 10¹³	0.00024 ETH
Avalanche	6 × 10¹⁵	0.006 AVAX
Polygon	6 × 10¹⁷	0.6 POL
BNB	1 × 10¹⁵	0.001 BNB
Linea	3 × 10¹³	0.00003 ETH
Plasma	6 × 10¹⁷	0.6 XPL
Ink	3 × 10¹³	0.00003 ETH

Wrapped token addresses

Each network has two wrapped-token addresses tracked in PaymentConfig:

• wrapped_native_token_address — the canonical wrapped version of the network’s native gas token (e.g. WETH on mainnet).
• wrapped_eth_address — the address of wrapped ETH on that network. On Ethereum and Arbitrum these are the same; on Avalanche, Polygon, BNB, Plasma, and others they differ.

Network	wrapped_native_token_address	wrapped_eth_address
Mainnet	0xc02aaa39b223fe8d0a0e5c4f27ead9083c756cc2 (WETH)	same
Gnosis	0xe91d153e0b41518a2ce8dd3d7944fa863463a97d (WXDAI)	0x6a023ccd1ff6f2045c3309768ead9e68f978f6e1
Arbitrum One	0x82af49447d8a07e3bd95bd0d56f35241523fbab1 (WETH)	same
Base	0x4200000000000000000000000000000000000006 (WETH)	same
Avalanche	0xB31f66AA3C1e785363F0875A1B74E27b85FD66c7 (WAVAX)	0x49D5c2BdFfac6CE2BFdB6640F4F80f226bc10bAB
Polygon	0x0d500B1d8E8eF31E21C99d1Db9A6444d3ADf1270 (WPOL)	0x7ceB23fD6bC0adD59E62ac25578270cFf1b9f619
BNB	0xbb4CdB9CBd36B01bD1cBaEBF2De08d9173bc095c (WBNB)	0x4db5a66e937a9f4473fa95b1caf1d1e1d62e29ea
Linea	0xe5d7c2a44ffddf6b295a15c148167daaaf5cf34f (WETH)	same
Plasma	0x6100e367285b01f48d07953803a2d8dca5d19873 (WXPL)	0x9895d81bb462a195b4922ed7de0e3acd007c32cb
Ink	0x4200000000000000000000000000000000000006 (WETH)	same

Orderbook database names

OrderbookConfig maps each network to a database name in the analytics PostgreSQL instance:

Network	network_db_name
Mainnet	mainnet
Gnosis	xdai
Arbitrum One	arbitrum-one
Base	base
Avalanche	avalanche
Polygon	polygon
BNB	bnb
Linea	linea
Plasma	plasma
Ink	ink

All networks use the dbt schema.

Nonce modifier system

Because COW token rewards for every network are paid from the same mainnet Safe, the pipeline must queue one transaction per network without nonce conflicts. It achieves this by adding a per-network integer offset to the current Safe nonce.

The default offsets are computed by reversing the Network enum declaration order and assigning indices starting at 0:

nonce_modifier_dict = {
    network: idx
    for idx, network in enumerate(reversed(list(Network)), start=0)
}

This produces the following default offsets (order may shift if new networks are added):

Network	Default nonce offset
MAINNET	9
BASE	8
ARBITRUM_ONE	7
BNB	6
AVALANCHE	5
POLYGON	4
GNOSIS	3
LINEA	2
PLASMA	1
INK	0

Override the automatic offset by setting the NONCE_MODIFIER environment variable when running the pipeline manually.

AccountingConfig.from_network() factory

AccountingConfig is the root configuration object that aggregates all sub-configs. It is constructed by calling the from_network() factory with the resolved Network enum value:

@staticmethod
def from_network(network: Network) -> AccountingConfig:
    return AccountingConfig(
        payment_config=PaymentConfig.from_network(network),
        orderbook_config=OrderbookConfig.from_network(network),
        dune_config=DuneConfig.from_network(network),
        node_config=NodeConfig.from_env(),
        reward_config=RewardConfig.from_network(network),
        protocol_fee_config=ProtocolFeeConfig.from_network(network),
        buffer_accounting_config=BufferAccountingConfig.from_network(),
        io_config=IOConfig.from_env(),
        overdraft_config=OverdraftConfig.from_network(network),
    )

The NETWORK environment variable is read inside IOConfig.from_env() and drives the Network enum resolution used by all other sub-config factories.