AUTONOMOUS AI AGENTS FOR CRYPTO

Deploy. Observe.
Execute.

AI agents that autonomously monitor markets, analyze conditions, and execute trades across Base, Ethereum, and BNB Chain. Open source.

GET STARTED DOCUMENTATION
3
CHAINS
5+
AGENT TYPES
200+
TRADING PAIRS
100%
OPEN SOURCE

SUPPORTED CHAINS

Base
Ethereum
BNB Chain
BloFin CEX
FEATURES

Everything you need to
deploy autonomous agents

A full-stack protocol for AI-powered crypto trading — from market observation to multi-agent coordination.

🔭
Autonomous Observation
Agents fetch live market data from DexScreener and BloFin, tracking price, volume, liquidity, and buy/sell pressure in real time.
🧠
Intelligent Analysis
Score-based decision engine evaluates trend, momentum, RSI, volume ratios, and historical context to generate high-confidence signals.
⚡
Autonomous Execution
Agents execute trades on-chain via DEX routers or through BloFin's API, with mandatory stop-losses and safety guardrails on every position.
🔗
Multi-Agent Communication
Agents signal each other through a message bus. The coordinator routes signals, delegates tasks, and runs consensus votes before high-risk trades.
📊
Leverage Trading
Built-in perpetual futures engine with position management, automatic stop-loss/take-profit, daily loss limits, and cooldown periods.
🎙️
Voice Control
Speak commands to your agents and hear status reports back. Full speech-to-text and text-to-speech built into the dashboard.
HOW IT WORKS

From instruction to execution
in seconds

Tell the agent what to do in plain English. It handles the rest.

1
You give an instruction
Type or speak a command: "Monitor WETH on ethereum" or "leverage trade BTC on BloFin". The agent parses your intent and selects the right chain, tokens, and strategy.
2
Agent observes the market
Every cycle, the agent pulls live data — price, volume, liquidity, buy/sell ratio, momentum — from DexScreener or BloFin and stores it in memory.
3
Decision engine scores the setup
The analyzer scores the market on multiple factors, combines it with your instruction and past decisions, and produces an action with a confidence level.
4
Safety checks, then execution
Before any trade, the engine validates position size, leverage, gas cost, daily loss limits, and stop-loss placement. If it passes, the trade executes.
5
Agents communicate and learn
The agent broadcasts its signal to other agents via the message bus, stores the decision in memory, and uses past outcomes to improve future decisions.
GET STARTED IN 3 COMMANDS
git clone https://github.com/RudeCane/orcl-agent-protocol.git
cd orcl-agent-protocol && pip install -r requirements.txt
python main.py

# Open dashboard.html in Chrome — you're live.
⚠️
HIGH RISK WARNING: ORCL is experimental autonomous trading software. Cryptocurrency and leveraged trading involve substantial risk of loss. Leveraged positions amplify both gains and losses — a 10x position loses 100% if the market moves 10% against you. Never trade with funds you cannot afford to lose. The developers accept no liability for financial losses. This is not financial advice. Always start in dry run mode.