@ekubo/sdk

Shared TypeScript math and protocol encoding utilities for Ekubo Protocol, supporting EVM and Starknet.

Installation

npm install @ekubo/sdk
# or
bun add @ekubo/sdk

Usage

All functions are pure and dependency-free. Exact protocol integers use native bigint values.

import {
  toSqrtRatio,
  computeStep,
  maxLiquidityForTokenAmounts,
} from "@ekubo/sdk";

// Convert a tick to a sqrt ratio for Starknet
const sqrtRatio = toSqrtRatio(0, "starknet");

// Compute a single swap step
const result = computeStep({
  sqrtRatio,
  liquidity: 1000000n,
  sqrtRatioLimit: toSqrtRatio(100, "starknet"),
  amount: 500n,
  isToken1: false,
  fee: 0n,
});
// result: { consumedAmount, calculatedAmount, sqrtRatioNext, feeAmount }

// Compute the maximum liquidity for a position given token amounts
const liquidity = maxLiquidityForTokenAmounts({
  sqrtPrice: sqrtRatio,
  sqrtPriceLower: toSqrtRatio(-100, "starknet"),
  sqrtPriceUpper: toSqrtRatio(100, "starknet"),
  amountBase: 1000n,
  amountQuote: 1000n,
});

API Reference

Types

Chain

type Chain = "evm" | "starknet";

Selects chain-specific tick and sqrt ratio bounds.

---

Constants

Name	Description
MAX_U128	Maximum value of a 128-bit unsigned integer
MAX_U256	Maximum value of a 256-bit unsigned integer
EVM_MIN_TICK	Minimum tick for EVM
EVM_MAX_TICK	Maximum tick for EVM
EVM_MIN_SQRT_RATIO	Minimum sqrt ratio for EVM
EVM_MAX_SQRT_RATIO	Maximum sqrt ratio for EVM
EVM_MAX_TICK_SPACING	Maximum tick spacing for EVM
STARKNET_MIN_TICK	Minimum tick for Starknet
STARKNET_MAX_TICK	Maximum tick for Starknet
STARKNET_MIN_SQRT_RATIO	Minimum sqrt ratio for Starknet
STARKNET_MAX_SQRT_RATIO	Maximum sqrt ratio for Starknet
STARKNET_MAX_TICK_SPACING	Maximum tick spacing for Starknet

---

Tick / Sqrt Ratio

toSqrtRatio(tick: number, chain: Chain): bigint

Converts a tick index to a fixed-point sqrt ratio for the given chain.

const sqrtRatio = toSqrtRatio(1000, "evm");

floatSqrtRatioToFixed(sqrtRatioFloat: bigint): bigint

Converts a compact 96-bit float sqrt ratio (as stored on-chain on Starknet mainnet) to the fixed-point representation used in math functions.

fixedSqrtRatioToFloat(sqrtRatio: bigint): bigint

Converts a fixed-point sqrt ratio to the compact float format expected on Starknet mainnet.

---

Token Amount Deltas

amount0Delta(sqrtRatioA, sqrtRatioB, liquidity, roundUp): bigint

Computes the amount of token0 (base token) between two sqrt ratios for a given liquidity.

amount1Delta(sqrtRatioA, sqrtRatioB, liquidity, roundUp): bigint

Computes the amount of token1 (quote token) between two sqrt ratios for a given liquidity.

---

Price / Sqrt Ratio Calculation

nextSqrtRatioFromAmount0(sqrtRatio, liquidity, amount0): bigint | null

Calculates the next sqrt ratio after trading a given amount of token0. Returns null if the result is out of range.

nextSqrtRatioFromAmount1(sqrtRatio, liquidity, amount1): bigint | null

Calculates the next sqrt ratio after trading a given amount of token1. Returns null if the result is out of range.

---

Swap

computeStep({ sqrtRatio, liquidity, sqrtRatioLimit, amount, isToken1, fee }): SwapResult

Computes a single step of a swap within a tick range.

• sqrtRatio – current sqrt ratio
• liquidity – active liquidity
• sqrtRatioLimit – sqrt ratio at which the swap stops (price limit)
• amount – signed amount to swap; negative means exact output
• isToken1 – true if the specified amount is in token1
• fee – fee as a fraction of 2^128

Returns:

{
  consumedAmount: bigint; // amount of the specified token consumed
  calculatedAmount: bigint; // amount of the other token
  sqrtRatioNext: bigint; // sqrt ratio after the step
  feeAmount: bigint; // fee collected
}

---

Liquidity

maxLiquidityForBaseToken({ sqrtPriceLower, sqrtPriceUpper, amount }): bigint

Returns the maximum liquidity achievable from a given amount of base token (token0) over a price range.

maxLiquidityForQuoteToken({ sqrtPriceLower, sqrtPriceUpper, amount }): bigint

Returns the maximum liquidity achievable from a given amount of quote token (token1) over a price range.

maxLiquidityForTokenAmounts({ sqrtPrice, sqrtPriceLower, sqrtPriceUpper, amountBase, amountQuote }): bigint

Returns the maximum liquidity achievable given both token amounts and the current price.

maxLiquidityForSpecifiedAmount({ sqrtPrice, sqrtPriceLower, sqrtPriceUpper, amount }): bigint

Returns the maximum liquidity for either a specified base or quote amount.

// Specify base amount
const liq = maxLiquidityForSpecifiedAmount({
  sqrtPrice,
  sqrtPriceLower,
  sqrtPriceUpper,
  amount: { base: 1000n },
});

// Specify quote amount
const liq2 = maxLiquidityForSpecifiedAmount({
  sqrtPrice,
  sqrtPriceLower,
  sqrtPriceUpper,
  amount: { quote: 1000n },
});

amountsFromSpecifiedAmount({ sqrtPrice, sqrtPriceLower, sqrtPriceUpper, amount }): { base, quote, maxLiquidity }

Given either a base or quote token amount, returns both token amounts and the maximum liquidity for the position.

liquidityToAmountBase({ sqrtPriceLower, sqrtPriceUpper, liquidity }): bigint

Converts a liquidity value to the required base token amount for a price range.

liquidityToAmountQuote({ sqrtPriceLower, sqrtPriceUpper, liquidity }): bigint

Converts a liquidity value to the required quote token amount for a price range.

---

TWAMM

calculateNextSqrtRatio(sqrtRatio, liquidity, token0SaleRate, token1SaleRate, timeElapsed, fee): bigint

Computes the next sqrt ratio for a TWAMM (Time-Weighted Average Market Maker) position after a given elapsed time.

---

Utilities

msb(x: bigint): number

Returns the position of the most significant bit of x.

---

EVM v3 Pool Keys

encodeEvmConcentratedPoolConfig and encodeEvmStableswapPoolConfig pack the 32-byte v3 config used by Core. decodeEvmPoolConfig returns the extension, exact bigint Q64 fee, pool discriminator, and type-specific parameters.

encodeEvmPoolKey returns the exact 96-byte ABI encoding hashed by Core. deriveEvmPoolId accepts a Keccak-256 function supplied by the consumer, which keeps the SDK dependency-free:

import { deriveEvmPoolId, encodeEvmConcentratedPoolConfig } from "@ekubo/sdk";
import { keccak256 } from "viem";

const config = encodeEvmConcentratedPoolConfig({
  fee: 2n ** 64n / 10_000n,
  tickSpacing: 200,
  extension: "0x0000000000000000000000000000000000000000",
});
const poolId = deriveEvmPoolId({ token0, token1, config }, keccak256);

Legacy position readers can use encodeEvmV2ConcentratedPoolConfig and encodeEvmV2StableswapPoolConfig; v2 configs deliberately omit the v3 pool type discriminator.

EVM Protocol Keys and Salted IDs

The SDK also owns the exact protocol encodings shared by applications and transaction preparers:

• encodeEvmAuctionConfig / decodeEvmAuctionConfig
• encodeEvmTwammOrderConfig / decodeEvmTwammOrderConfig
• calculateEvmTwammMaxSaleRate
• deriveEvmAuctionTokenId, deriveEvmTwammOrderTokenId, and deriveEvmVeTokenId
• deriveEvmIndexedSalt

Hashing functions accept a caller-supplied Keccak-256 implementation, keeping the SDK dependency-free while using the same ABI preimages everywhere.

MEVCapture v2.1 (main variant)

MevCaptureV21MainPool quotes a concentrated pool on the current EVM Core behind the MEVCapture v2.1 main extension. It reproduces every CORE.swap call the extension makes: the toward call, charged away calls cut one step short of the next rate boundary or initialized tick, bridges onto initialized ticks, and the per-call surcharge passed to accumulateAsFees (taken from the output on exact input, added to the input on exact output). The first-touch anchor update and post-swap refresh run in simulateSwap and quote.

import { MevCaptureV21MainPool } from "@ekubo/sdk";

const pool = new MevCaptureV21MainPool({
  poolFee, // the pool key's Q0.64 fee
  tickSpacing,
  config, // the extension's getConfig tuple
  state: { core: { sqrtRatio, tick, liquidity }, anchor }, // anchor: getAnchor
  sortedTicks, // every initialized tick the swap can reach
});
const quote = pool.quote({
  isToken1: false,
  amount: 10n ** 18n,
  now: blockTimestamp,
});

Quotes are exact only for a complete initialized-tick snapshot, because the extension cuts calls at Core's bitmap tick events. The replay in test/mevCaptureV21Main.test.ts checks all 21 contract vectors (266 Core calls) with no tolerance.

---

Development

This project uses Bun.

# Install dependencies
bun install

# Run tests
bun test

# Build
bun run build

License

MIT