TradeClaw Documentation
TradeClaw is MIT-licensed, self-hostable trading-signal software with a rule-based analysis engine, modeled paper trading, optional Telegram delivery, and a plugin system. Follow the documented Docker Compose and environment setup for your infrastructure.
What is TradeClaw?
TradeClaw generates candidate signal records with a multi-indicator rule engine for assets and timeframes supported by the configured data source. Its confluence score describes weighted indicator agreement; it is not a calibrated probability of profit.
The entire TA engine is pure TypeScript — no external indicator libraries, no black boxes. You can read every formula, fork it, and extend it with the plugin system.
Architecture
Next.js App (apps/web)
All routes, API endpoints, dashboard UI, and React components.
TA Engine (apps/web/app/lib/ta-engine.ts)
Pure-math RSI, MACD, EMA, Bollinger Bands, Stochastic. Zero dependencies.
Plugin System (apps/web/lib/plugin-system.ts)
Sandboxed JS execution for custom indicators (VWAP, ATR, OBV, Ichimoku).
Signal Scanner (Dockerfile.scanner)
Background worker that polls prices, runs TA, and emits signals on a configurable interval.
TimescaleDB + Redis
PostgreSQL with time-series extensions for signal history; Redis for rate limiting.
Documentation
Getting Started
Core Features
Integrations
Quick Start
The fastest way to run TradeClaw is Docker Compose. From the repo root:
# 1. Clone the repository
git clone https://github.com/naimkatiman/tradeclaw
cd tradeclaw
# 2. Configure environment
cp apps/web/.env.example apps/web/.env.local
# 3. Start all services
docker compose up -d
# App runs at localhost:3000
See the Installation guide for Railway, Vercel, and bare-metal deployment options.