Polymarket News-Driven Trading Bot

Phase 2: Live paper trading with real market data, state persistence, analytics, and reporting.

Architecture

News/Twitter ──→ Collectors ──→ SQLite ──→ Signal Analyzer ──→ Paper Engine
                    │                          │  ↕ fallback         │
                    │                    LLM (primary)          Risk Manager
                    │                    Keyword (fallback)         │
                    │                                        Settlement Check
                    ↓                                              ↓
              Analytics ←────────────────────────────── Feishu Reporter
              (Sharpe, Calibration, Export)              (Daily + Alerts)

Quick Start

pip install -r requirements.txt
cp .env.example .env  # fill in LLM_API_URL, LLM_API_KEY, WEBHOOK_URL
python -m src.main --config config/default.yaml

Docker

cp .env.example .env  # fill in secrets
docker-compose up -d

Systemd

sudo cp polymarket-bot.service /etc/systemd/system/
sudo systemctl enable --now polymarket-bot

Components

Module	Purpose
Collectors	News RSS (4 feeds) + Twitter (3 accounts via fxtwitter)
Signal Analyzer	LLM semantic analysis (DeepSeek V3.2) + keyword fallback
Paper Engine	Virtual portfolio, state persistence, trade journal
Risk Manager	Quarter Kelly sizing, daily loss limit, drawdown kill switch
Analytics	Sharpe ratio, confidence calibration, CSV/JSON export
Reporter	Feishu daily reports + critical alerts

Database (SQLite)

Table	Purpose
news_items	Collected RSS articles
tweets	Collected tweets
signals	Generated signals (dedup by hash)
trades	All paper trades
trade_journal	Full trade context (news sources, LLM reasoning)
portfolio_state	Single-row portfolio snapshot
daily_metrics	Per-day performance metrics

Phase Roadmap

• Phase 1 ✅: Paper trading scaffold
• Phase 2 ✅: State persistence, analytics, Feishu reports, Docker
• Phase 3: Small capital live trading