This project implements a modular, rule-based yield agent that performs DeFi strategy optimization across lending protocols, AMM pools, and delta-neutral positions. It is designed for stepwise automation, with a roadmap to evolve into a fully autonomous AI agent.

yield-agent-main/
├── scripts/
│   ├── lending_borrowing/         # Tracking APRs on lending/borrowwing protocols (live + historical)
│   ├── volume_tracking/           # LP fee estimation via tick-level liquidity and volume
│   ├── liquidity_management/      # Manage LP positions: initialize, monitor, rebalance
│   └── delta_neutral/             # Delta neutral system for hedging LP exposure
├── utils/                         # Additional analysis scripts
├── abis/                          # ABIs for Ethereum contracts
└── README.md                      # ← You are here

Folder: scripts/lending_borrowing/

Tracks lending and borrowing rates across DeFi protocols to identify capital-efficient opportunities for arbitraging, yield farming, collateralized strategies, or leveraged LP positions.

Currently supports:

• Aave V3 on Ethereum, Arbitrum, and Base
• Tokens: WETH, USDC, WBTC

Planned additions:

• Morpho
• Compound V3
• Other lending markets

Scripts:

• fetch_live_aave_rates.js – Fetches live APRs from Aave V3 smart contracts
• fetch_past_aave_rates.js – Loads historical APR data from AaveScan and stores it in PostgreSQL and Excel
• query_aave_rates.js – Queries the local rate database for analytics and strategy selection
• monitor_rates.js – Compares lending/borrowing rates across protocols and chains to detect arbitrage or optimization opportunities (work in progress)

💡 This module will evolve into a unified rate oracle that compares real-time lending and borrowing yields across multiple DeFi platforms.

Folder: scripts/volume_tracking/

Estimates yield from swap fees by analyzing tick-level liquidity and volume in Uniswap V3 pools. This module provides granular insight into fee-generating zones and helps evaluate the performance of liquidity positions.

Scripts:

• 1_agg_info_pool.js – Fetches pool metadata (token info, decimals, tick spacing, fee tier)
• 2_liquidity_per_tick.js – Retrieves gross/net liquidity per tick from the pool
• 3_active_liquidity_per_tick.js – Determines the active tick range and filters relevant liquidity
• 4_swapEvents_per_tick.js – Parses recent swaps and assigns them to tick bins
• 5_merge_swaps_and_liquidity.js – Merges swap volume and liquidity data per tick
• 6_fees_per_tick.js – Computes estimated fees using volumeUSD * feeTier

Planned extensions:

• 🧠 Fee-based APR estimation per tick or range
• 📈 LP strategy simulation and backtesting
• 🔁 Multi-day or rolling window fee tracking
• 🌐 Support for other AMMs:
  • PancakeSwap V3 (BNB Chain, Base)
  • Orca Whirlpools (Solana)
  • Uniswap V4 and Hooks

💡 This module will evolve into a cross-DEX fee oracle, enabling optimal LP placement, strategy backtesting, and real-time APR estimation.

Folder: scripts/liquidity_management/

Tracks and manages active liquidity positions in Uniswap V3 pools, with automated rebalancing. The goal is to maintain optimal liquidity placement based on price movement, volume trends, and fee yield projections.

Scripts:

• initialize_LP_position.js – Adds new LP positions to selected tick ranges (currently being integrated)
• monitor_LP_position.js – Monitors if a position is in range and triggers alerts or rebalance logic (currently being integrated)
• overview_LP_positon.js – Displays LP metrics such as current tick, liquidity, unclaimed fees, and position health (currently being integrated)

Planned extensions:

• 🧠 Auto-rebalancing logic triggered by price deviation or fee drop-off
• 📊 Historical performance tracking per LP position
• 🔄 Strategy switching (e.g. range shift, withdrawal, reinvestment)
• 🌐 Support for additional AMMs:
  • PancakeSwap V3
  • Uniswap V4 (Hooks)
  • Orca Whirlpools (via SDK on Solana)

💡 This module will evolve into a fully autonomous LP management engine that monitors, rebalances, and rotates LP positions based on real-time market conditions.

Folder: scripts/delta_neutral/

Implements a modular intent-based execution system to hedge LP exposure or take directional positions based on predefined trading logic. Combines on-chain intent storage with an off-chain execution relayer.

Currently live on:

• Hyperliquid (Arbitrum) – supports directional intents such as long/short USDC/ETH

Core components:

• 🧾 Smart contract: IntentStorage.sol – Stores trading intents on-chain using enums and minimal storage
• 🤖 Off-chain relayer: relayer.py – Listens for emitted events and executes/cancels trades based on on-chain state
• 🛠 CLI tool: createIntent.py – Command-line utility to create, update, delete, or check intents

📦 How it works:

1. Create trading intent → Store on-chain (createIntent.py)
2. Event emitted → Off-chain relayer listens and executes trade (relayer.py)

📂 Key files:

• createIntent.py
• relayer.py
• IntentStorage.sol

Planned extensions:

• 🌐 Cross-chain deployment (e.g. Base, Optimism, Solana)
• 🔁 Support for more platforms:
  • Drift (Solana)
  • Vertex (Arbitrum)
  • Aevo (Ethereum)
• 🔒 Multi-intent and failover execution logic
• 🧠 Agent-based dynamic intent creation based on market conditions

💡 This system forms the foundation for an autonomous hedging layer, enabling delta-neutral strategies in response to LP risk, price volatility, or yield shifts across chains.

This project is the foundation for a self-optimizing yield agent, with future improvements:

• ✅ Rule-based execution (current)
• 🔄 Dynamic triggers from live price feeds
• 🧠 Agent decision engine (planned)
• 📈 Real-time dashboards (future frontend)

To get started, review the individual README.md files inside each folder for module-specific instructions, dependencies, and environment setup.

# Navigate into each subfolder and follow the README instructions:
cd scripts/lending_borrowing/
cat README.md

cd scripts/volume_tracking/
cat README.md

cd scripts/liquidity_management/
cat README.md

cd scripts/delta_neutral/
cat readme.MD  # note: lowercase extension

# Most modules require .env variables, RPC endpoints, and/or PostgreSQL setup

Built by hwdeboer1977
Part of a broader mission to create modular, intelligent agents for DeFi.