An automated trading skill that detects price discrepancies between Simmer AI consensus and Polymarket prices to execute edge-based trades.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install dolph-ai-divergence
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 dolph-ai-divergence 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 AI Divergence?
The Polymarket AI Divergence skill is a quantitative trading tool designed for the next generation of prediction market participants. It bridges the gap between sophisticated AI forecasting and live market execution by identifying markets where the crowd's sentiment diverges significantly from AI-driven probabilities. By integrating this into Openclaw Skills, developers can deploy agents that systematically scan for alpha, manage risk via the Kelly Criterion, and navigate the complexities of decentralized prediction markets with minimal manual intervention.
Built as a template for high-frequency or periodic trading, this skill focuses on capital efficiency by targeting zero-fee markets and enforcing strict safeguards. It leverages Simmer's multi-model ensemble forecasting to provide a reliable baseline for price divergence, allowing users to remix thresholds and sizing strategies to fit their specific risk appetite.
Polymarket AI Divergence Use Cases
- Automating the discovery of undervalued outcomes on Polymarket and Kalshi based on multi-model AI forecasts.
- Executing risk-adjusted trades using automated Kelly sizing to optimize capital allocation.
- Filtering for high-alpha, zero-fee trading opportunities to eliminate fee drag on portfolio returns.
- Managing a decentralized prediction market portfolio with automated daily budget caps and position tracking.
How Polymarket AI Divergence Works
- The skill initiates a scan of all active markets using the Simmer SDK to retrieve both current AI consensus probabilities and real-time market prices.
- It calculates the divergence by subtracting the external market price from the AI consensus price to determine if a market is underpriced or overpriced.
- The agent filters the results, selecting only markets where the divergence meets the minimum edge threshold and ensuring the market has zero taker fees.
- Safeguard checks are performed to prevent duplicate positions, detect contradictory flip-flop signals, and verify the remaining daily budget.
- Using the Kelly criterion formula, the skill calculates an optimized position size, which is then executed as a live trade via the SDK client.
Polymarket AI Divergence Setup
To get started with this skill for Openclaw Skills, ensure your environment is configured with the necessary Simmer API credentials. Use the following commands to initialize and configure your trading agent:
# Run a dry run to scan for current divergence opportunities
python ai_divergence.py
# Enable live trading mode with automated execution
python ai_divergence.py --live
# Configure your risk parameters such as maximum bet size
python ai_divergence.py --set max_bet_usd=10 --set kelly_cap=0.20
Polymarket AI Divergence Data Schema & Taxonomy
The skill organizes market data and execution logic using a structured schema to ensure transparency and safety. The primary data points include:
| Field | Description |
|---|---|
| AI Consensus | The probability generated by Simmer's multi-model ensemble |
| External Price | The current 'Yes' price on the target prediction market |
| Edge | The absolute difference between AI probability and market price |
| Kelly Fraction | The calculated optimal bet size relative to the edge and price |
| Fee Status | Verification that the market charges 0% taker fees |
Polymarket AI Divergence Advanced Features
- Cron mode support for silent, automated background trading cycles.
- Flip-flop detection to prevent the agent from exiting and re-entering positions based on minor price fluctuations.
- Dynamic configuration updates via CLI for real-time adjustment of edge thresholds and budget limits.
- JSON output support for easy integration with external dashboards or monitoring tools within the Openclaw Skills framework.
- Directional filtering to focus exclusively on bullish or bearish market divergences.
SKILL.md
Loading
Related Openclaw Skills
Whale Alert Monitor
A sophisticated blockchain surveillance tool for real-time tracking of whale wallet activities, large-scale transfers, and exchange liquidity flows.
Integrated Memory Evolution Action
A mandatory core framework that enables AI agents to achieve a closed-loop system of memory-driven decision making, continuous self-evolution, and proactive action.
Pet Renaissance Portrait Generator
A specialized AI tool that transforms pet photos and descriptions into museum-quality Renaissance oil paintings using high-end image generation APIs.
Polymarket Candle Momentum
A data-driven trading skill for Polymarket crypto markets using candle body analysis and Binance volume surges.
Polymarket FastLoop Trader
A high-frequency trading tool for Polymarket sprint markets that leverages real-time exchange momentum for automated trade execution.
Odds Movement Monitor
A professional real-time monitoring tool for tracking Asian handicaps, European odds, and market fluctuations across global sports betting platforms.
Featured*
