Architecture

High-Level Flow

Market Data (Hyperliquid API)
        │
        ▼
┌─────────────────────────┐
│   OHLCV + Order Book    │◄──── Coinglass On-Chain Data
│   Multi-Timeframe       │      (OI, Funding, LSR, Liquidations)
└─────────┬───────────────┘
          │
          ▼
┌─────────────────────────┐
│  Market Regime Detector │ ──► 9 States (Strong Trend, Ranging, Volatile...)
│  + Adaptive Parameters  │
└─────────┬───────────────┘
          │
          ▼
┌─────────────────────────┐
│  Strategy Engine         │ ──► 13+ Strategies run in parallel
│  (Confluence-based)      │     Each generates scored signals
└─────────┬───────────────┘
          │
          ▼
┌─────────────────────────┐
│  Signal Quality Pipeline │
│  ┌───────────────┐      │
│  │ Tier 2 Quality│      │ ──► Quality filters, R:R validation
│  │ ML Filter     │      │ ──► Gradient Boosted decision
│  │ Correlation   │      │ ──► Portfolio exposure check
│  │ Tiered Conf.  │      │ ──► Confidence erosion protection
│  │ Entry Timing  │      │ ──► Candle close, chase detection
│  └───────────────┘      │
└─────────┬───────────────┘
          │
          ▼
┌─────────────────────────┐
│  AI Validation           │ ──► Claude AI evaluates thesis
│  (ECTRA AI v2.0)         │     APPROVE / ADJUST / REJECT
└─────────┬───────────────┘
          │
          ▼
┌─────────────────────────┐
│  Risk Manager            │ ──► Position sizing, leverage calc
│  + Feature Gate          │     Tier-based access control
└─────────┬───────────────┘
          │
          ▼
┌─────────────────────────┐
│  Execution Engine        │ ──► Hyperliquid order placement
│  (execute.py)            │     5-sig-fig price rounding
└─────────┬───────────────┘
          │
          ▼
┌─────────────────────────┐
│  Position Management     │
│  ┌───────────────┐      │
│  │ ATR Trailing   │      │ ──► Volatility-based trail
│  │ Smart S/R     │      │ ──► Ichimoku + Pivot S/R trail
│  │ CG Trailing   │      │ ──► Coinglass-intelligence trail
│  │ Thesis Monitor│      │ ──► Continuous thesis validation
│  └───────────────┘      │
└─────────┬───────────────┘
          │
          ▼
┌─────────────────────────┐
│  Telegram + Persistence  │ ──► Alerts, commands, trade journal
│  + Trading Metrics       │     SQLite state persistence
└─────────────────────────┘

Module Map (64 Files)

Core Engine

Module	Purpose
main.py	Master orchestrator — scan loop, signal routing, position monitoring
execute.py	Hyperliquid API wrapper — orders, positions, market data, MTF OHLCV
config_v9.py	Configuration — strategy modes, risk parameters, credentials (XOR-encrypted)
risk_manager.py	Position sizing, leverage calculation, trailing stop phases
master_system.py	Multi-strategy signal aggregation and unified execution

Strategy Modules

Module	Strategy	Category
strat_smc.py	Smart Money V2	SMC / Order Blocks / FVG
strat_orderflow.py	Order Flow	CVD Divergence / Volume
strat_of_mtf.py	Order Flow MTF ⭐	Flagship MTF strategy
strat_scalp.py	Enhanced Scalping	Multi-condition scoring
strat_session.py	Session Liquidity	Session high/low sweeps
strat_pattern.py	Chart Pattern	Classical pattern recognition
strat_sniper.py	Scalp Sniper	High-frequency OB/FVG/Sweep
strat_smc_mtf.py	Smart Money MTF	SMC with HTF bias
strat_confluent.py	Confluent	Multi-strategy agreement
strat_diversified.py	4 Diversified Strategies	MR, Momentum, Funding, LiqCascade

Analysis & Intelligence

Module	Purpose
regime.py	9-state market regime detection + adaptive parameter system
order_flow.py	Real-time WebSocket order flow + OHLCV-derived flow analysis
zones.py	Premium/Discount zones + MTF pattern confirmation
structure.py	Swing point detection, market structure (HH/HL/LH/LL)
indicators.py	Core technical indicator calculations
institutional.py	Liquidation level estimation
correlation.py	Cross-asset correlation and portfolio exposure
microstructure.py	Spread, depth, liquidity regime analysis
session_filter.py	Session-based filtering + regime TP/SL adjustment
coinglass.py	Coinglass helper (internal strategy use)
coinglass_integration.py	Full Coinglass API integration (v2.0 Institutional Pro)

AI & Machine Learning

Module	Purpose
ectra_ai_v2.py	AI signal validation, news sentiment, performance learning
ectra_ai_chat.py	Interactive AI trading assistant (GUI dialog)
ectra_news_monitor.py	Real-time news monitoring with AI sentiment
ectra_strategy_builder.py	AI-powered custom strategy creation
ectra_indicator_database.py	1,870-line indicator reference for AI context
ml_filter.py	ML feature extraction + gradient boosted signal filter

Risk & Quality

Module	Purpose
tier2_quality.py	Signal quality enhancement filters
tier3_risk.py	Tier 3 advanced risk management
tier4_analyzers.py	News, breadth, volume profile, smart money analysis
tiered_confidence.py	Confidence erosion protection (soft/hard classification)
thesis_monitor.py	Post-entry thesis validity monitoring
entry_timing.py	Candle close, chase detection, sweep confirmation
mr_time_decay.py	Mean Reversion-specific exponential time decay
trading_metrics.py	Profit factor, R:R tracking, performance analytics

Trailing Stop Systems

Module	Purpose
atr_trailing.py	ATR-based trailing stop (5 modes)
smart_trailing.py	S/R + Ichimoku Cloud trailing (6 phases)
smart_cg_trailing.py	Coinglass-intelligence trailing (6 phases)

Interface & Infrastructure

Module	Purpose
gui_app.py	PyQt6 desktop GUI with live dashboard
settings_dialog.py	GUI settings configuration dialog
telegram_bot.py	Telegram alerts, commands, license admin
vectra_persistence.py	SQLite trade journal and state persistence
vectra_backtest_engine.py	Institutional backtesting (Monte Carlo, Walk-Forward)
vectra_backtest_panel.py	GUI backtesting panel
license_validator.py	License key validation
license_generator.py	License key generation (admin)
snipe_entry_v2.py	Precision wick capture module
zone_entry_module.py	Reusable zone entry + pending order system
analysis.py	Technical analysis utilities
verify_setup.py	System verification on first run
build_exe.py	Windows executable builder

Data Flow Philosophy

VECTRA follows a strict Pipeline Architecture where data flows through increasingly selective filters.

1

Raw Data

Market data from exchange + on-chain data from Coinglass.

2

Regime Classification

Determine what type of market we are in.

3

Signal Generation

Multiple strategies generate scored candidates.

4

Quality Pipeline

Signals pass through 5+ quality filters.

5

AI Validation

Independent AI review of the trade thesis.

6

Risk Check

Portfolio-level exposure and sizing validation.

7

Execution

Precise order placement with tick-size compliance.

8

Active Management

Triple trailing + thesis monitoring until exit.

Each layer can reject the signal. A trade only executes if it survives every filter. This multi-gate approach dramatically reduces the number of trades while ensuring each executed trade has the highest possible probability of success.

Defensive Design Principles

Every external call (exchange API, Coinglass API, AI API, WebSocket) is wrapped in try/except blocks with graceful fallback behavior. If Coinglass is unavailable, the bot continues with technical analysis only. If the AI API fails, the signal proceeds with a confidence penalty. If the WebSocket disconnects, order flow falls back to OHLCV-derived estimates.

The system is designed to never crash from external failures and to always degrade gracefully.