ditus/uniswap_auto_clp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
main
uniswap_auto_clp/florida/doc/SECURITY_AND_OPTIMIZATION.md

5.4 KiB
Raw Permalink Blame History

Security & Optimization Protocols

This document details the safety mechanisms, entry guardrails, and optimization strategies implemented in the Uniswap Auto CLP & Hedger system.

1. CLP Manager: Entry & Position Safety

The clp_manager.py module is responsible for opening and managing Concentrated Liquidity Positions (CLP). It implements strict checks to ensure positions are only opened under favorable conditions.

A. Oracle Guard Rail (Anti-Manipulation)

Goal: Prevent entering a pool that is actively being manipulated (Flash Loans) or has momentarily de-pegged.

  • Mechanism: Before any calculation, the bot fetches the Real-Time Mid Price directly from the Hyperliquid API.
  • Logic:
    • Pool Price is calculated from the on-chain sqrtPriceX96.
    • Oracle Price is fetched from Hyperliquid (https://api.hyperliquid.xyz/info).
    • Rule: If abs(Pool - Oracle) / Oracle > 0.25%, the bot ABORTS the entry.
  • Benefit: Protects against entering at a "fake" price which would lead to instant arbitrage loss (LVR).

B. Stale Tick Protection (Volatility Guard)

Goal: Prevent entering a position if the price moves significantly during the transaction setup phase (e.g., while swapping tokens).

  • Mechanism:
    1. The bot calculates the target tick range at T_0.
    2. It performs necessary token swaps (e.g., USDC -> WETH) which takes time (T_1).
    3. Critical Check: Just before minting (T_2), it re-fetches the current pool tick.
  • Rule: If abs(Tick_T0 - Tick_T2) > 13 ticks (approx 0.13%), the bot ABORTS the mint transaction.
  • Benefit: Ensures the range is centered on the actual execution price, not an outdated one.

C. Post-Mint Accuracy

Goal: Ensure the "Entry Price" recorded for the Hedger is 100% accurate.

  • Mechanism: Immediately after the Mint transaction is confirmed on-chain, the bot fetches the pool state one more time.
  • Benefit: Captures the exact price impact of the user's own liquidity insertion, preventing discrepancies in the Hedger's PnL calculations.

D. Safe Entry Zones (AUTO Mode)

Goal: Only enter when mean reversion is statistically likely.

  • Mechanism:
    • Bollinger Bands (BB): Price must be inside the 12h BB.
    • Moving Average (MA): The MA88 must also be inside the 12h BB.
  • Benefit: Avoids opening positions during breakout trends or extreme volatility expansion.

E. Manual Override (Force Mode)

Goal: Allow operator intervention for testing or recovery.

  • Command: python clp_manager.py --force <width> (e.g., 0.95).
  • Behavior: Bypasses Oracle, BB, and MA checks for the first position only. Automatically disables itself after one successful mint.

2. CLP Hedger: Risk Management

The clp_hedger.py module manages the Delta-Neutral hedge on Hyperliquid.

A. Emergency Edge Closure (Stop-Loss)

Goal: Prevent indefinite hedging losses if the price breaks out of the LP range violently.

  • Trigger: If Price >= Range Upper Bound.
  • Action: Immediately CLOSE the short hedge position.
  • Benefit: Stops the strategy from "selling low and buying high" (hedging) when the LP position is already out of range and effectively essentially 100% stablecoin (impermanent loss realized).

B. Hysteresis Reset

Goal: Prevent "whipsaw" losses (opening/closing repeatedly) if the price hovers exactly at the edge.

  • Logic: Once an Emergency Closure triggers, the hedge does not re-open immediately if the price dips slightly back in.
  • Reset Condition: Price must drop back to a "Safe Zone" (e.g., 75% of the range width or a specific buffer below the edge).
  • Benefit: Filters out noise at the range boundaries.

C. Asymmetric Compensation (EAC)

Goal: Reduce the "Buy High/Sell Low" churn near the edges of the range.

  • Mechanism: The target hedge delta is adjusted (reduced) as the price approaches the boundaries.
  • Logic: Target Hedge = Pool Delta * (1 - Proximity_Factor).
  • Benefit: Softens the impact of entering/exiting the range, preserving capital.

D. Fishing Orders (Maker Rebates)

Goal: Reduce execution costs.

  • Mechanism: Instead of market dumping (Taker fee ~0.035%), the bot places Limit Orders (Maker) slightly away from the spread.
  • Logic: If the price moves favorably, the order fills, earning a rebate (or paying 0 fee). If not filled within TIMEOUT, it falls back to a Taker order if the delta drift is critical.

3. Future Improvements (Roadmap)

A. WebSocket Integration

  • Current: Polling REST API every 30-300s (or 1s for guard rails).
  • Upgrade: Implement a persistent WebSocket connection to Hyperliquid and the EVM RPC.
  • Benefit: Sub-100ms reaction times to volatility events.

B. Multi-Chain Arbitrage Check

  • Current: Checks Hyperliquid vs Pool.
  • Upgrade: Check Binance/Coinbase prices.
  • Benefit: Detects on-chain lag before it happens (CEX leads DEX).

C. Dynamic Fee Optimization

  • Current: Fixed POOL_FEE.
  • Upgrade: Automatically switch between 0.05% and 0.01% pools based on volatility and volume metrics.

D. Smart Rebalance (Inventory Management)

  • Current: Swaps surplus tokens using 1inch/Universal Router.
  • Upgrade: Use CowSwap or CoW intents for MEV-protected rebalancing of large inventory imbalances.