Oracle sniper command:

export PYTHONPATH=$PYTHONPATH:. && python3 src/scripts/sync_market_data.py --no-sentiment

TrueData Downloads

To download today's data for all existing symbols: python3 scripts/download_today_batch.py

To download data for the last 3 days for all symbols: python3 scripts/download_today_batch.py --days 3

To download just today's data for a single specific symbol: python3 scripts/tick_downloader.py --symbol RELIANCE --days 0

python3 scripts/download_today_batch.py --exchange BSE --days 2

Dhan daily ohlc :

python3 scripts/download_historical_ohlc.py --days 1

DELETE FROM intraday_ticks WHERE timestamp < DATE_SUB(NOW(), INTERVAL 2 DAY);

---

To implement your strategies, you can subscribe to the Redis channel market_depth or poll the keys depth:.

Every packet is a JSON string with the following structure:

1. Data Fields Field Type Description symbol string The ticker symbol (e.g., "SBIN"). ltp float Last Traded Price. imbalance float Order book imbalance (Level 1). Range: -1.0 (Sell heavy) to 1.0 (Buy heavy). vwap float Real-time calculated Intraday VWAP. vqs_score float Momentum score based on tick-by-tick direction. total_bid_qty int Combined quantity of all 20 levels of the bid-side book. total_ask_qty int Combined quantity of all 20 levels of the ask-side book. bid_pct / ask_pct float Percentage of the total book volume on each side. buy_vol / sell_vol int Aggregated aggressive buyer/seller volume since start of day. bids / asks list Top 5 price/quantity pairs: [[price, qty], ...] bids_20 / asks_20 list Full 20 levels (Only present for the primary 50 focus symbols).
2. Example Structure json { "symbol": "OLAELEC", "ltp": 28.21, "imbalance": 0.05, "vwap": 28.18, "vqs_score": 0.1245, "total_bid_qty": 3124500, "total_ask_qty": 2845000, "bid_pct": 52.34, "ask_pct": 47.66, "buy_vol": 150000, "sell_vol": 120000, "timestamp": "2026-02-18T20:10:21.952", "source": "DHAN", "exchange_bridge": "Dhan20", "bids": [[28.21, 4500], [28.20, 10000], [28.19, 5000], [28.18, 1200], [28.17, 3000]], "asks": [[28.22, 4200], [28.23, 8000], [28.24, 15000], [28.25, 2000], [28.26, 5000]], "bids_20": [ ... 20 levels ... ], "asks_20": [ ... 20 levels ... ] }
3. Usage Tips Momentum Strategy: Use the vqs_score. If > 0.5, the price is aggressively ticking up. Liquidity Strategy: Compare total_bid_qty vs total_ask_qty for high-resolution sentiment. Scale: You can monitor up to 500 of these packets per second across your local Redis instance.

redis replay

---

Start Shadow Redis: redis-server --port 6380 Run Replay: python3 -m src.scripts.parquet_replay --date 2026-02-06 --port 6380 --speed 5.0 Run Signal Generator: REDIS_PORT=6380 python3 -m src.services.signal_generator

redis channels:

Institutional Activity (Iceberg Alerts) Channel: icebergs Frequency: On detection (Asynchronous) Purpose: Immediate notification when large hidden orders are absorbing retail liquidity. Field Type Description

symbol String Target symbol. ltp Float Price where absorption occurred. actiontobetaken String BUY (Hidden buyer) or SELL (Hidden seller).

time String ISO Timestamp of the peak volume tick. Example Message:

json { "symbol": "RELIANCE", "ltp": 2942.15, "actiontobetaken": "BUY", "time": "2026-02-22T11:45:02" }

Main Depth & Orderflow Hub Channel: market_depth:LIVE Frequency: Every Price/Depth tick (High Frequency) Purpose: Real-time signal generation, PCR sentiment, and Orderflow metrics. Field Type Description

symbol String The base NSE symbol (e.g., "RELIANCE"). ltp Float Last Traded Price from NSE. imbalance Float Orderbook bias (-1.0 to 1.0). >0 means more Bids. vqs_score Float Price momentum (-1.0 to 1.0). Directional velocity. oi Float Real-time Open Interest (mapped from active Future). pcr_oi Float Put-Call Ratio (based on full Option Chain OI). max_pain Float Price strike where option writers lose the least. bids / asks List Top 5 price/qty pairs. bids_20 / asks_20 List Full 20-level depth (for priority focus symbols). Example Message:

json { "symbol": "SBIN", "ltp": 785.45, "imbalance": 0.35, "vqs_score": 0.12, "vwap": 784.10, "oi": 5420000.0, "pcr_oi": 1.15, "max_pain": 780.0, "bid_pct": 62.5, "ask_pct": 37.5, "buy_vol": 1500, "sell_vol": 200, "timestamp": "2026-02-22T10:15:30.123", "bids": [[785.40, 500], [785.35, 1200]], "asks": [[785.50, 400], [785.55, 900]] }

1. Restart the OrderFlow Orchestrator Since the Momentum Squeeze and VWAP Deviation logic is now built directly into the Orchestrator, you just need to restart it.

If you are running it natively on your Mac (recommended for lowest latency), run this in a terminal:

bash export PYTHONPATH=$PYTHONPATH:. python3 src/services/orderflow_orchestrator.py 2. Start the Background Market Scanner The Volume Surge Scanner is a new standalone service. You should start it in a separate terminal window so it can continuously monitor the database for surges across all stocks:

bash export PYTHONPATH=$PYTHONPATH:. python3 src/services/vol_surge_scanner.py

1. Kill the old orchestrator

pkill -f "src.services.orderflow_orchestrator"

2. Start in "Momentum Only" mode

export ENABLED_SIGNALS=MOMENTUM_SQUEEZE export PYTHONPATH=$PYTHONPATH:. python3 -m src.services.orderflow_orchestrator > logs_momentum_only.txt 2>&1 &