kaccy-core

Core business logic for the Kaccy Protocol — a personal Future Output Token (FOT) platform. This crate houses all domain models, pricing engines, trading infrastructure, DeFi primitives, ML/analytics pipelines, and shared utilities that power the Kaccy exchange.

Scale (v0.2.0): 210 Rust source files · ~90,548 lines · 4,663 public items · 1,640 tests passing.

Features

Models

Full domain-model layer covering every entity in the protocol: User, Token, Order, Trade, Balance, Governance, KYC, Leaderboard, Payment, Position, Reputation, Vesting, Airdrop, and Notification — each with Serde serialization and SQLx-compatible types.

Pricing

Seven production-ready bonding curve implementations backed by high-precision rust_decimal arithmetic, plus a dynamic fee engine, fee optimizer, price oracle, staking rewards, liquidity-mining incentives, and buyback automation.

Trading

A complete order-routing stack: limit/market order book, market maker, AMM, arbitrage engine, flash-loan executor, MEV detection and mitigation, HFT metrics, smart order routing, backtesting harness, batch auctions, conditional orders, options pricing, VPIN calculation, and commit-reveal order submission.

DeFi

Margin trading, a lending protocol, perpetual futures, concentrated-liquidity AMM, liquid staking, yield farming, synthetic assets, P2P lending, intent-based trading, JIT liquidity provisioning, distributed consensus, cross-margin accounting, and auto-compounding vaults.

ML / Analytics

Time-series forecasting (ARIMA, GARCH, Prophet), ensemble methods, sentiment analysis, GPU-accelerated inference, online learning, reinforcement learning, feature engineering, and model-persistence utilities — all implemented in pure Rust without Python bindings.

Utils

A broad utility layer including risk management, portfolio optimization, on-chain analytics, multi-level caching, compliance checks, encryption, KYC/AML screening, cross-chain bridge adapters, database optimization (sharding, replicas), fuzz testing, load testing, memory profiling (TrackingAllocator, MemoryProfiler, MemoryGuard), structured metrics, distributed tracing, webhook dispatch, and threshold encryption.

Architecture

Bonding Curves

All seven curves implement the BondingCurve trait and expose spot_price, buy_price (integral over supply range), and sell_price methods. SquareRoot and Logarithmic curves use closed-form O(1) antiderivatives.

Installation

[dependencies]
kaccy-core = "0.2.0"

Quick Start

Pricing a bonding curve

use kaccy_core::pricing::{BondingCurve, LinearBondingCurve};
use rust_decimal_macros::dec;

let curve = LinearBondingCurve::new(
    dec!(0.0001),   // base price (BTC)
    dec!(0.00001),  // slope: price increment per token
);

// Spot price at supply 100
let spot = curve.spot_price(dec!(100));

// Cost to purchase 10 tokens starting at supply 100 (integral)
let cost = curve.buy_price(dec!(100), dec!(10));
println!("spot={spot}  cost={cost}");

Submitting an order

use kaccy_core::models::{Order, OrderSide, OrderType};
use kaccy_core::trading::OrderBook;
use rust_decimal_macros::dec;
use uuid::Uuid;

let mut book = OrderBook::new();

let order = Order::new(
    Uuid::new_v4(),          // user_id
    Uuid::new_v4(),          // token_id
    OrderSide::Buy,
    OrderType::Limit,
    dec!(10),                // quantity
    Some(dec!(0.0015)),      // limit price (BTC)
);

let fills = book.submit(order)?;

Testing

Property tests verify mathematical invariants for every bonding curve: monotonicity, non-negativity, round-trip buy/sell consistency, and supply conservation across arbitrary (supply, quantity) pairs generated by proptest.

Performance

• SquareRoot and Logarithmic curves use closed-form antiderivatives for O(1) integral computation instead of numerical summation.
• SIMD pricing paths are enabled when target-feature includes AVX2 or NEON, reducing batch-price calculation latency for HFT and batch-auction workloads.
• The TrackingAllocator / MemoryProfiler pair in utils::memory_profile lets callers measure per-operation allocation cost in tests and benchmarks.

Dependencies

License

Copyright COOLJAPAN OU (Team Kitasan). All rights reserved.