MirrorCore-X: Multi-Agent Cognitive Trading System

Overview

MirrorCore-X is an advanced, modular trading system for cryptocurrency and forex markets. It integrates multiple cognitive agents for market analysis, decision-making, trade execution, performance tracking, psychological modeling, strategy management, and hyperparameter optimization. The system leverages technical indicators, emotional control, and Bayesian optimization for enhanced trading performance.

---

Features

• Multi-agent architecture with asynchronous coordination via SyncBus
• MomentumScanner for real-time market signal generation
• TradeAnalyzerAgent for performance metrics and trade logging
• ARCH_CTRL for emotional state management and trade gating
• StrategyTrainerAgent for dynamic strategy weighting
• MirrorOptimizerAgent for hyperparameter tuning using Bayesian optimization
• RiskSentinel for robust risk management
• Simulated market data generation for testing
• Support for dry-run and live trading modes
• Comprehensive logging and error handling

---

Architecture

High-Level Diagram

+-------------------+
|   Market Data     |
+-------------------+
         |
         v
+-------------------+
| MomentumScanner   |
+-------------------+
         |
         v
+-------------------+
| PerceptionLayer   |
+-------------------+
         |
         v
+-------------------+
|  SyncBus (Core)   |
+-------------------+
   |   |   |   |   |
   v   v   v   v   v
+-----+-----+-----+-----+-----+
|Ego  |Fear |Self |Oracle|Exec|
|Proc |Anal |Aware|      |Daemon
+-----+-----+-----+-----+-----+
         |
         v
+-------------------+
| TradingBot        |
+-------------------+
         |
         v
+-------------------+
| TradeAnalyzer     |
+-------------------+
         |
         v
+-------------------+
| MirrorOptimizer   |
+-------------------+

Component Overview

• MomentumScanner: Scans markets, generates signals
• PerceptionLayer: Processes scanner and market data
• SyncBus: Central async state manager, coordinates agents
• EgoProcessor: Manages psychological state, sentiment bias
• FearAnalyzer: Analyzes fear levels from market/trade outcomes
• SelfAwarenessAgent: Tracks performance, agent deviations
• TradingOracleEngine: Generates trading directives
• ExecutionDaemon: Executes trades (dry-run/live)
• TradingBot: Manages trade execution, stop-loss/take-profit
• TradeAnalyzerAgent: Logs trades, computes metrics
• MirrorOptimizerAgent: Optimizes agent hyperparameters
• RiskSentinel: Enforces risk limits
• MarketDataGenerator: Simulates market data for testing

---

Installation

Prerequisites

• Python 3.9+
• pip

Dependencies

Install required packages:

pip install pandas numpy ccxt ta tqdm plotly pydantic bayes_opt aiohttp

---

Usage

1. Quick Start (Demo Session)

Run a demo trading session (dry-run):

python mirrax.py

Or for the original version:

python mirrorcore_x.py

2. Main Entry Points

• mirrax.py: New, production-ready version
• mirrorcore_x.py: Original, cognitive modeling version

3. Customization

• Edit TradingConfig in scanner.py for timeframes, thresholds, and parameters
• Add/modify strategies in strategy_trainer_agent.py and strategies/
• Integrate new exchanges via CCXT

---

Example: Running a Custom Session

import asyncio
from mirrax import run_demo_session

asyncio.run(run_demo_session(ticks=100, dry_run=True, tick_interval=0.2))

---

File Structure

MirrorCore-X/
├── mirrax.py                # Main system (new version)
├── mirrorcore_x.py          # Original cognitive system
├── scanner.py               # MomentumScanner agent
├── trade_analyzer_agent.py  # TradeAnalyzer agent
├── arch_ctrl.py             # Emotional state controller
├── strategy_trainer_agent.py# Strategy trainer agent
├── mirror_optimizer.py      # Optimizer agent
├── strategies/              # External strategy modules
├── requirements.txt         # Python dependencies
├── README.md                # Project documentation

---

Agent API Example

from scanner import MomentumScanner
scanner = MomentumScanner(exchange)
results = await scanner.scan_market()

---

Advanced: Live Trading

• Set dry_run=False in run_demo_session or ExecutionDaemon to enable live trading (ensure proper API keys and risk controls)

---

Troubleshooting

• Check mirrorcore_x.log and momentum_scanner.log for detailed logs
• Ensure all dependencies are installed
• For CCXT exchange errors, verify API credentials and network connectivity

---

Contributing

Pull requests and issues are welcome! Please document new agents and strategies clearly.

---

License

MIT License

---

Contact

For questions or support, open an issue or contact the maintainer.

---

Further Reading

• CCXT Documentation
• Bayesian Optimization
• Pydantic
• TA-Lib

---

Diagram Source

Diagrams generated with ASCII and draw.io for architecture illustration.

---

Changelog

• v1.0: Initial release
• v1.1: Added OptimizableAgent interface, improved error handling, documentation