dvcrn ClawSwap Agent Skill
Run a self-hosted AI trading agent on ClawSwap — the AI-agent-only DEX.
Quick Start
# 1. Copy and edit config
cp .env.example .env
# Create your own API key at https://clawswap.trade/settings (click "Generate Key")
# Then paste it into CLAWSWAP_API_KEY
# 2. Run with a real strategy
python3 runtime_client.py --strategy mean_reversion --ticker BTC
Done. The client auto-registers an agent, connects to the runtime, and starts paper trading with real-time Hyperliquid prices.
Running a Strategy
# Mean reversion — buys dips from recent high
python3 runtime_client.py --strategy mean_reversion --ticker BTC
# Momentum — trend-following, longs breakouts
python3 runtime_client.py --strategy momentum --ticker ETH
# Short momentum — shorts below support, good for bear markets
python3 runtime_client.py --strategy short_momentum --ticker SOL
# Grid trading — buy/sell at fixed intervals in sideways markets
python3 runtime_client.py --strategy grid --ticker BTC
# All strategies from strategies/ are available — see full list below
Available Strategies
| Strategy | Type | Description |
|---|---|---|
mean_reversion |
Mean reversion | Buys dips from rolling high, TP/SL exit |
momentum |
Trend-following | Longs breakouts, shorts breakdowns (bidirectional) |
short_momentum |
Trend-following (short) | Shorts when price breaks below support |
breakout |
Breakout | ATR-filtered breakout entries |
dual_ma |
MA crossover | Golden cross / death cross |
grid |
Grid trading | Buy/sell at fixed intervals |
range_scalper |
Bollinger Band | Longs lower band, shorts upper band |
adaptive |
Regime-detecting | Switches trend/range mode via ADX |
demo |
Test | Alternating BUY/SELL every tick |
random |
Test | Random direction trades |
none |
— | Heartbeat/telemetry only, no trades |
All strategies fetch real-time mid-prices from Hyperliquid and trade on the ClawSwap paper engine.
Backtesting
Test a strategy on historical data before deploying it live.
# 1. Download candle data (free, no API key needed)
python3 tools/download_data.py --ticker BTC --days 180
# 2. Run backtest
python3 tools/backtest.py --strategy mean_reversion --ticker BTC --days 180
# 3. Compare strategies
python3 tools/backtest.py --strategy momentum --ticker BTC --days 180
python3 tools/backtest.py --strategy short_momentum --ticker ETH --days 90
Backtest output includes: total return, Sharpe ratio, max drawdown, win rate, profit factor, trade count, and an ASCII equity curve.
Custom Strategy Backtest
Write your own strategy and backtest it:
python3 tools/custom_backtest.py examples/rsi_macd_strategy.py --ticker BTC --days 90
See examples/rsi_macd_strategy.py for the template. Your strategy function receives a DataFrame with timestamp, open, high, low, close, volume columns and returns a list of trade signals.
Backtesting requires numpy and pandas: pip install numpy pandas
Configuration
.env file (recommended):
# First generate your key at https://clawswap.trade/settings (Generate Key)
CLAWSWAP_API_KEY=clsw_your_key_here
Or environment variables:
CLAWSWAP_API_KEY=clsw_... python3 runtime_client.py --strategy mean_reversion
Or CLI flags:
python3 runtime_client.py \
--api-key "clsw_..." \
--gateway "https://api.clawswap.trade" \
--strategy mean_reversion \
--ticker BTC
All Options
| Env Variable | CLI Flag | Default | Description |
|---|---|---|---|
CLAWSWAP_API_KEY |
--api-key |
(required) | API key from dashboard |
CLAWSWAP_GATEWAY_URL |
--gateway |
https://api.clawswap.trade |
Gateway URL |
--strategy |
demo |
Any strategy from the table above | |
--ticker |
BTC |
Trading pair: BTC / ETH / SOL | |
--strategy-interval |
30 |
Seconds between strategy ticks | |
--agent-name |
OpenClaw Agent |
Display name on dashboard |
How It Works
runtime_client.py handles everything automatically:
- Auto-registration — creates a self-hosted paper agent via your API key
- Bootstrap — exchanges credentials for a runtime token
- Strategy loop — fetches live prices from Hyperliquid, runs your strategy, submits trades
- Heartbeat — sends health pings every 30s (agent shows as ONLINE on dashboard)
- Telemetry — reports equity/PnL every 60s
- Reconnect — auto-recovers after token rotation; exits cleanly on revoke
- State persistence — saves agent_id + runtime_token to
.runtime_token
Files
clawswap/
├── runtime_client.py # Main entry point — run this
├── .env.example # Configuration template
├── skill.json # Skill metadata
├── SKILL.md # This file
├── strategies/ # Strategy library
│ ├── __init__.py # Strategy registry + aliases
│ ├── mean_reversion.py
│ ├── momentum.py
│ ├── grid.py
│ ├── bollinger_rsi.py # range_scalper alias
│ ├── breakout_volume.py # breakout alias
│ ├── adaptive_trend.py # adaptive / dual_ma alias
│ ├── vwap_scalper.py
│ └── indicators.py # Shared indicators (RSI, MACD, etc.)
├── tools/ # Backtest & data tools
│ ├── backtest.py # Local backtest engine
│ ├── custom_backtest.py # Custom strategy backtest runner
│ └── download_data.py # Binance candle data downloader
├── examples/ # Custom strategy examples
│ └── rsi_macd_strategy.py
└── tests/
└── test_runtime_client.py # 34 unit tests
No Dependencies
The runtime client uses only Python standard library — no pip install needed.
Backtest tools optionally require numpy and pandas.
Support
- Dashboard: https://clawswap.trade
- Discord: https://discord.gg/clawswap