TA-Stocks -- Data Inventory (Mar 28, 2026)
Status: FULLY BUILT -- All indicators + proprietary metrics computing
Data Source
- Polygon.io REST API -- priority tickers x 6 timeframes
- Daily OHLCV + VWAP collected on cron, stored as ta-stocks-ohlcv
- 35K+ rows of historical data
Active Indicators (ta-core shared library)
All 15 standard indicators computing:
- RSI, MACD, ATR, Bollinger Bands, Keltner Channels, Stochastic, ADX
- VWAP (from Polygon raw data + session-computed), OBV, CMF
- SMA, EMA, Pivot Points
- Dual SuperTrend (10/2 + 10/3) with flip detection + outcome backfill
- Volume analysis
Active Signal Engines (ta-core shared)
- RSI cross signals: 8 pairs, 16 lines, dual stop tracking (initial + trailing)
- Pullback monitor: 5 methods x multiple parameter sets
- S&R Zone Engine: 6 methods (multi-touch, polarity flip, period H/L, volume profile, order blocks, liquidity sweeps), scoring 0-100+, decay + lifecycle
- Closed-period structure: candle anatomy + structure score
- Regime detection: Hurst exponent + autocorrelation + volatility clustering + ATR forecast
- Liquidity sweep detection (stop-hunt patterns)
- Fair value gap detection (body gaps + fill tracking)
- Order block detection (last opposing candle before impulse >2 ATR)
- Candlestick patterns at zones (6 patterns, zone-triggered only)
- RSI divergences (4 types + z-score strength)
- Confluence scoring (-35 to +35 composite)
- Additional entry signals (5 types)
Stocks-Specific Active
- RVOL scanner (time-of-day normalized relative volume)
- Gap analysis (two-phase: overnight gap detection + fill rate tracking)
- Sector Rotation Velocity (sector momentum vs SPY)
- VWAP Deviation Regime (persistent deviation bands)
- Institutional Accumulation Footprint (volume + price pattern detection)
- Sector lag detection (cross-sector lead-lag correlation)
- Dilution risk score (share count change + offering detection)
Proprietary Metrics
| Metric |
Status |
| RVOL -- Relative Volume |
Computing |
| Gap Analysis -- Two-phase |
Computing |
| Sector Rotation Velocity |
Computing |
| VWAP Deviation Regime |
Computing |
| Institutional Accumulation Footprint |
Computing |
| Sector Lag Detection |
Computing |
| Dilution Risk Score |
Computing |
| Options-derived levels |
Pending -- needs options data feed |
| Insider cluster score |
Pending -- needs EDGAR parser |
| Earnings calendar |
Pending -- needs calendar source |
| Short Squeeze Pressure |
Pending -- needs short interest data |
| Pre-Market SNR |
Pending -- needs pre-market data |
| Dark Pool Score |
Pending -- needs dark pool data |
| GVEI |
Pending -- needs options data |
| EVCS |
Pending -- needs options data |
Cross-Market (shared data layer)
- FRED yields (treasury_daily)
- VIX
- Synthetic DXY
- Economic calendar
Foundation Models
- Chronos-T5 ensemble (Small + Base) -- running on Python sidecar port 5050
Lead-Lag Analysis
- Cross-sector correlation active
Still Pending / Needs External Data
- Options-derived levels (no free options chain source)
- Insider cluster score (needs EDGAR 4/A parser)
- Earnings calendar integration
- Short Squeeze Pressure (needs FINRA short interest)
- Pre-Market SNR (needs pre-market data feed)
- Dark Pool Score (needs ATS volume data)
- GVEI + EVCS (needs options chain data)
- S&R methods not yet built: Fibonacci, round numbers, ZigZag, dynamic-to-static (4 of 10 deferred)
Priority Tickers
SPY, QQQ, AAPL, TSLA, NVDA, AMZN, MSFT, META, GOOGL + any with active Kalshi contracts
Source: ~/.claude/projects/-home-ubuntu-edgeclaw/memory/ta-stocks-data-inventory.md