rule-based-trading

A local service for rule-based automated trading with Kite Connect (Zerodha). Define exit rules with take-profit, stop-loss, and trailing stops - the engine monitors your positions and executes exits automatically.

Features

• Multi-user support - Each user has their own broker credentials and rules
• Encrypted credentials - Broker API keys are encrypted in the database
• Rule-based exits - Define TP/SL conditions via REST API
• Trailing stops - Automatically adjust stop-loss as price moves in your favor
• Time conditions - Set trading hours and auto square-off times
• REST API - Full control via HTTP endpoints
• Database storage - Rules persist in PostgreSQL

Quick Start

1. Install Dependencies

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

2. Configure Environment

cp .env.example .env
# Edit .env with your configuration

SECRET_KEY=your-secret-key-change-this-in-production
DATABASE_URL=postgresql+asyncpg://user:password@localhost:5432/trading
REDIS_URL=redis://localhost:6379/0

3. Start the Server

uvicorn main:app --reload

4. Create an Account

Register via the API:

curl -X POST http://localhost:8000/auth/register \
  -H "Content-Type: application/json" \
  -d '{"email": "user@example.com", "password": "secure_password"}'

5. Connect Your Broker

Store your Kite Connect credentials:

curl -X POST http://localhost:8000/broker/connect \
  -H "Authorization: Bearer <your_token>" \
  -H "Content-Type: application/json" \
  -d '{"api_key": "your_kite_api_key", "api_secret": "your_kite_api_secret"}'

6. Define Trading Rules

Create rules via the API:

curl -X POST http://localhost:8000/rules \
  -H "Authorization: Bearer <your_token>" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "SENSEX Call Option",
    "symbol_pattern": "SENSEX*",
    "exchange": "BFO",
    "take_profit": {
      "enabled": true,
      "condition_type": "relative",
      "target": 100
    },
    "stop_loss": {
      "enabled": true,
      "condition_type": "relative",
      "stop": 40
    }
  }'

7. Start the Rule Engine

curl -X POST http://localhost:8000/engine/start \
  -H "Authorization: Bearer <your_token>"

Rule Configuration

Condition Types

Type	Description	Example
relative	Points from entry	target: 100 = entry + 100
percentage	Percent of entry	target: 15 = entry × 1.15
absolute	Fixed price level	target: 800 = exit at 800

Position Types

• LONG: Bought position → TP triggers on price UP, SL on price DOWN
• SHORT: Sold position → TP triggers on price DOWN, SL on price UP

Trailing Stops

{
  "stop_loss": {
    "enabled": true,
    "condition_type": "relative",
    "stop": 50,
    "trail": true,
    "trail_step": 50
  }
}

Time Conditions

{
  "time_conditions": {
    "start_time": "09:15",
    "end_time": "15:15",
    "square_off_time": "15:20",
    "active_days": [0, 1, 2, 3, 4]
  }
}

API Endpoints

Authentication

Endpoint	Method	Description
/auth/register	POST	Create a new account
/auth/login	POST	Login and get JWT token
/auth/status	GET	Check authentication status

Broker

Endpoint	Method	Description
/broker/connect	POST	Store broker credentials
/broker/oauth	GET	Initiate OAuth flow
/broker/status	GET	Check broker connection

Engine Control

Endpoint	Method	Description
/engine/start	POST	Start rule evaluation
/engine/stop	POST	Stop rule evaluation
/engine/status	GET	Get engine status

Rules Management

Endpoint	Method	Description
/rules	GET	List all rules
/rules	POST	Create new rule
/rules/{id}	GET	Get rule details
/rules/{id}	PUT	Update rule
/rules/{id}	DELETE	Delete rule

Positions & Trades

Endpoint	Method	Description
/positions	GET	Get current positions
/trades/active	GET	Get active trades

Development

Run Tests

pytest tests/ -v

Format Code

black .

Docker

docker-compose up -d

License

MIT