Algo Trade

Automated NIFTY50 Options Trading bot based on Global sentiment, institutional flows, and real-time technical indicators.

---

🚀 Latest: Browser-based Full Automation

The current version is a fully browser-based automated trading system for Nifty weekly options. It runs as a React dashboard backed by a Cloudflare Worker proxy and a D1 database for persistence. Order execution and tracking are fully automated via the Upstox API.

Key Features:

• 5-Layer Scoring Engine: Combines macro sentiments, technical indicators, and institutional flow into a single unified bullish/bearish score.
• Web Dashboard: Modern UI available at app/client to control the bot, configure strategies, view signals, and manage positions.
• D1 Persistence: Safely persists broker accounts and strategy configuration via a Cloudflare Worker backed by a SQLite D1 database.
• State Machine Polling: Robust state management (IDLE ➔ RUNNING ➔ ORDERED ➔ RUNNING/STOPPED) polling every 60s.
• Automated Exit Logic: Triggered automatically on target profit/loss, indicator/macro reversals, or index breadth flips.

For complete details on the strategy rules, see docs/v5-strategy.md.

---

🛠️ Setup Guide

Web Client & Worker Setup

To run the browser dashboard locally and deploy its backend:

1. Navigate to the client directory:

   cd app/client

2. Install the required dependencies:

   yarn install

3. Apply D1 migrations to set up the persistence database locally:

   yarn wrangler d1 migrations apply algo-trade-paper --local

4. Run the development server:

   yarn dev

5. (Optional) Apply migrations to your remote Cloudflare database and deploy the Worker:

   yarn wrangler d1 migrations apply algo-trade-paper --remote
   yarn deploy

For detailed client setup and migration steps, refer to app/client/README.md and docs/client-persistence.md.

---

V1 - V4 Console Application Setup (Legacy CLI)

If you wish to run the legacy CLI/console versions (V1–V4):

1. Navigate to the core directory:

   cd app/core

2. Install dependencies:

   yarn install # or npm install

3. Create a .env file and configure your broker credentials. Use .env.sample as a template.
4. Compile the TypeScript code:

   tsc

5. Start the console job:

   yarn start # or npm run start

Console Logs Preview:

---

📐 Architecture & Scoring Engine

Browser Tab (React + Vite)
│
├─ useStrategyBot.ts          ← State machine, 60s polling loop
│   ├─ fetchMarket()          ← Candles + Option Chain + Macro signals
│   └─ fetchMarketSentiment() ← VIX, FII, DII, PCR, Max Pain + synthetic MMI/PE/IV
│
├─ strategyEngine.ts          ← 5-layer scoring → FinalSignal
│   ├─ scoreBullish()         ← BUY CE score (max ~27 pts)
│   └─ scoreBearish()         ← BUY PE score (max ~26 pts)
│
├─ Cloudflare Worker          ← CORS proxy for all external APIs
│   ├─ Upstox API             ← Candles, option chain, FII/DII, VIX, place/exit orders
│   └─ MoneyControl API       ← Global indices technical ratings
│
└─ UI (strategies page)
    ├─ MarketSetupPanel        ← VIX, FII %, Nifty PE, MMI
    ├─ InstitutionalPanel      ← MMI gauge, FII scores
    ├─ BreadthPanel            ← A/D ratio, PCR zone, Straddle IV
    ├─ IndicatorsPanel         ← 6 technical indicator cards
    ├─ ScorePanel              ← Bull/bear score bars + final signal
    ├─ BotControls             ← Start/Stop, position card, countdown
    └─ StrategyConfig          ← Config form with localStorage

The 5 Scoring Layers

Layer	Source	Signal & Description
L0: Hard Stops	Upstox VIX	Blocks trading completely if VIX is out of bounds (< 10 or > 25). Nifty PE is synthetic and penalises score only.
L1: Macro	MoneyControl + Upstox breadth	Evaluates global index sentiment (Dow, Nikkei, Hang Seng, FTSE, etc.), Advance/Decline ratios, and Put-Call Ratio (PCR).
L2: Technicals	Upstox 1-min candles	Real-time indicators: EMA crossover (10/42), ADX, RSI, Stochastic, Bollinger Bands, and ATR.
L3: Institutional	Upstox FII/DII API + synthetic computation	Institutional flow: MMI (synthetic), FII Long/Short ratio (bull contrarian + bear momentum), Net Positioning, Straddle IV vs VIX.
L4: Confluence Gate	Unified Evaluator	Enforces minimum score gap (bull vs bear) and overall score threshold to generate entry signals.

---

⚙️ Strategy Configurations (Editable in UI)

These parameters can be customized dynamically from the dashboard UI:

• Confidence Thresholds: Customize minimum scores for both moderate and strong signals.
• Profit & Loss Limits: Set trailing or absolute targets for automatic position exits.
• Max Trades Per Day: Protect your account by limiting over-trading.
• Last Entry Time: Restrict bot entry signals past a set time (e.g., 14:30 IST).
• Strike Offset (OTM Skip): Configurable offset to purchase out-of-the-money options (default: 3 strikes OTM).

---

📜 Release History

• V5 (Latest): React Dashboard with Cloudflare Worker proxy, D1 schema storage, and multi-layer scoring engine.
• V4 (In Dev / Console): Integrates technical indicators directly from Upstox 1-min candles (EMA crossover, ADX, RSI, Stochastic, Bollinger Bands, ATR).
• V3: Consolidated macro sentiment fetches from MoneyControl and NSE India, placing orders via Upstox.
• V2: Scraped Investing.com for global trends and 1-min indicators, executed trades via Finvasia.
• V1: Basic Investing.com scraper with 5-min/1-min trend checks and ATR/RSI, trading via Finvasia.

---

🤝 Collaboration & Contact

We welcome contributions! If you would like to collaborate:

• Connect with me on LinkedIn: itaravin
• Submit issues or feature requests: GitHub Issues

Trade Logs (stored in AWS for console apps):