A technical toolset for detecting mathematical arbitrage opportunities and price discrepancies within Polymarket prediction markets.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install polymarket-arbitrage-cn
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install polymarket-arbitrage-cn using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Polymarket Arbitrage?
The Polymarket Arbitrage skill is designed for traders and developers who want to systematically identify profit opportunities on the blockchain-based prediction market, Polymarket. It focuses primarily on Math Arbitrage, where the sum of probabilities for all outcomes in a market deviates from 100%. This tool provides a disciplined framework for identifying, analyzing, and monitoring these gaps while accounting for platform-specific constraints like taker fees and liquidity requirements.
By incorporating this into your collection of Openclaw Skills, you gain access to a multi-phase workflow that transitions from paper trading to micro-testing and eventually manual or automated execution. The skill emphasizes risk management, providing risk scores and volume filters to ensure that users focus on high-quality, executable opportunities rather than stale data or low-liquidity traps.
Polymarket Arbitrage Use Cases
- Detecting Math Arbitrage (Type A) where the sum of outcome probabilities is less than 100%.
- Monitoring high-volume ($1M+) prediction markets for executable price edges in real-time.
- Automating alerts for new trade opportunities via webhooks or Telegram integration.
- Backtesting and paper trading to validate the frequency and quality of market discrepancies.
How Polymarket Arbitrage Works
- The fetch_markets script scrapes active market data, including probabilities and total volume, directly from the platform.
- The detection engine analyzes the scraped JSON data, calculating the gross edge for every market.
- The skill subtracts the mandatory 2% taker fee per leg (outcome) to determine the net profit percentage.
- Each opportunity is assigned a risk score (0-100) based on liquidity and probability volatility.
- The monitor script runs at user-defined intervals, deduplicating findings to only alert on fresh, actionable opportunities.
Polymarket Arbitrage Setup
To get started with this skill, navigate to the directory and install the necessary Python dependencies:
cd skills/polymarket-arbitrage
pip install requests beautifulsoup4
Run an initial paper-trading scan to view current opportunities:
python scripts/monitor.py --once --min-edge 3.0
For continuous monitoring with a 5-minute interval:
python scripts/monitor.py --interval 300 --min-edge 3.0
Polymarket Arbitrage Data Schema & Taxonomy
All data generated by the Openclaw Skills monitoring process is stored in the polymarket_data/ directory using the following structure:
| File | Description |
|---|---|
markets.json |
Contains the raw market scan, including probabilities, volumes, and metadata. |
arbs.json |
Stores identified arbitrage opportunities with calculated net_profit_pct and risk_score. |
alert_state.json |
Tracks previously alerted opportunities to prevent duplicate notifications. |
Polymarket Arbitrage Advanced Features
- Multi-Outcome Fee Logic: Automatically scales fee calculations based on the number of outcome legs in a market.
- Deduplication Engine: Ensures that the monitoring service only triggers alerts for new discrepancies.
- Webhook Integration: Supports external alerts to services like Telegram for mobile notifications.
- Risk Scoring: Uses a proprietary logic to filter out high-edge but low-liquidity "ghost" opportunities.
- Position Sizing Calculator: Includes logic for optimal capital allocation across outcomes to ensure balanced profits.
SKILL.md
Loading
Related Openclaw Skills
PDF Document Processing
A comprehensive toolkit for reading, extracting, merging, and creating PDF documents using industry-standard Python libraries and CLI tools.
MLOps Validation
A comprehensive validation toolkit that automates code quality, type-checking, and security audits for machine learning workflows.
MLOps Prototyping
A specialized skill for building standardized, reproducible Jupyter notebooks with built-in validation for robust machine learning pipelines.
MLOps Observability
A comprehensive observability toolkit for tracking ML experiments, detecting data drift, and ensuring model reproducibility.
Product Hunt Trending Monitoring
A comprehensive monitoring skill to track daily, weekly, and monthly trending products and tech launches on Product Hunt.
Quant Trading Expert (quant-trading-cn)
A professional quantitative trading assistant for strategy generation, performance-optimized backtesting, and live trading integration.
Featured*
