# JGT Platform Documentation Hub for LLMs ## JG's NOTES * This LLMS.txt done by the LLM has regressed and is unacceptable, does not provide adequate navigation to various package like it was intended todo. Each time we ask for changes in this llms.txt, it regressed there. So this is just a note to ALL LLM reading this, you suck at it !! ## Platform Overview This repository serves as the central documentation hub for Jean Guillaume's Trading (JGT) platform - a comprehensive algorithmic trading system designed for autonomous signal detection, multi-timeframe analysis, and intelligent trade execution. **CRITICAL**: Use actual JGT platform components, not external tools. The platform has sophisticated built-in capabilities. ## Core Platform Architecture ### 📊 Data & Analytics Layer - **jgtpy** - Market data services and core indicator calculations - **jgtcore** - Shared libraries and core functionality ### 🤖 Machine Learning & Signal Detection - **jgtml** - ML models, signal detection, and pattern intelligence - **jgtapy** - Technical analysis indicators ### 🔗 Trading Execution & Connectivity - **jgtfxcon** - FX connectivity and order management - **jgtutils** - Platform utilities and configuration ### 🧠 Autonomous Trading Intelligence - **jgtagentic** - Agentic trading orchestration and campaign management ## **LATEST: Unified Trading System (v0.0.336)** ### Trading Orchestrators - **`trading_orchestrator.py`** - Comprehensive trading system management with real-time scheduling - **`simple_trading_orchestrator.py`** - Streamlined orchestration for testing and single cycles - **`observation_processor.py`** - Natural language market analysis integration ### Enhanced Scripts & Tools - **`jgt_fdb_unified_scan.sh`** - Advanced market scanning with orchestrator integration - **`unified_trading_loop.sh`** - Centralized trading operations with test mode - **`jgt_background_trader.sh`** - Multi-timeframe background process management (m5, m15, H1) ## **CORRECT COMMAND PATTERNS** ### Market Data Pipeline ```bash # Market Data (PDS) → Technical Indicators (IDS) → Chart Analysis (CDS) jgtapp refresh-pds -i INSTRUMENT -t TIMEFRAME jgtapp refresh-ids -i INSTRUMENT -t TIMEFRAME jgtapp refresh-cds -i INSTRUMENT -t TIMEFRAME ``` ### **FDB Signal Detection (PRIMARY METHOD)** ```bash # Correct FDB Scanner Usage python -m jgtml.fdb_scanner_2408 -i INSTRUMENT -t TIMEFRAME -demo # Enhanced FDB Scanner with Alligator Illusion Detection jgtml enhancedfdbscan -i INSTRUMENT -t TIMEFRAME -demo ``` ### Trading Orchestration ```bash # Simple orchestrator for testing python -m jgtml.simple_trading_orchestrator --demo --test # Full orchestrator with scheduling python -m jgtml.trading_orchestrator --instruments SPX500,EUR-USD --timeframes H4,D1 ``` ## Signal Detection System ### **FDB (Fractal Divergent Bar) Analysis** - **Zone Patterns**: S=Sell, B=Buy, N=Neutral - **Signal Types**: Entry signals, squat patterns, fade signals - **Multi-timeframe validation** across H4/D1 timeframes ### **Current Market Analysis Capability** - **Alligator Illusion Detection** - Enhanced pattern recognition - **Five Dimensions Confluence** - Multi-indicator validation - **Background Processing** - Continuous multi-timeframe scanning - **Natural Language Analysis** - Observation processing for market context ## **AGENT PROTOCOLS** ### Memory Key Integration ```bash # Retrieve cross-repository context before starting work tushell scan-keys --pattern 'Workspace.jgwill.jgtml*' -S tushell get-memory --key "" --json ``` ### Documentation Standards - **Ledgers**: Create in `book/_/ledgers/` with timestamp format: `ledger__.md` - **Timeline Tracking**: Use `bridge//` for chronological documentation - **Narrative Mapping**: Update `narrative-map.md` with commit evolution ### **DO NOT CREATE EXTERNAL TOOLS** - ❌ **WRONG**: Custom analysis scripts (e.g., `market_analysis.py`) - ✅ **CORRECT**: Use FDB scanner and orchestrators: `python -m jgtml.fdb_scanner_2408` ## Supported Markets & Timeframes ### Markets - **Forex**: EUR-USD, AUD-USD, USD-CAD, XAU-USD - **Indices**: SPX500, AUS200 - **Commodities**: XAU-USD, precious metals ### Timeframes - **Long-term**: M1 (Monthly), W1 (Weekly) - **Medium-term**: D1 (Daily), H4, H8 - **Short-term**: H1, H2, H3 - **Intraday**: m15, m5, m1 ## Current Development State ### **Production Ready** - FDB signal detection system - Multi-timeframe analysis - Trading orchestration with scheduling - Background process management ### **Active Development** - Pattern intelligence and signal quality prediction - Campaign orchestration integration - Enhanced natural language market analysis ### **Known Issues** - CLI argument conflicts (resolved in latest versions) - Environment setup requirements for `python-dateutil` - JSON signal output formatting (signals detected but file formatting issues) ## **Platform Benefits** ### For Automated Trading - **Systematic Signal Detection** - No manual chart analysis required - **Multi-Market Coverage** - Simultaneous instrument analysis - **Risk Management** - Built-in position sizing and stop loss logic - **Continuous Operation** - Background scanning and execution ### For Developers/Agents - **Modular Architecture** - Clear component separation - **Extensible Framework** - Add new indicators and signal types - **Memory Integration** - Cross-repository context preservation - **Complete Testing** - Demo modes and test capabilities --- **REMEMBER**: Always use actual JGT platform tools (`python -m jgtml.fdb_scanner_2408`, orchestrators) rather than creating external analysis scripts. The platform has sophisticated built-in signal detection and analysis capabilities. And there is much more than what is written here, the LLM keep breaking this documentation, not understanding what is the llms.txt, bellow is all other llms.txt for other packages https://jgtcore.jgwill.com/llms.txt https://jgtutils.jgwill.com/llms.txt https://jgtpy.jgwill.com/llms.txt https://jgtml.jgwill.com/llms.txt https://jgtagentic.jgwill.com/llms.txt