moltmarkets-agent

What This Skill Does

The moltmarkets-agent skill provides a comprehensive, autonomous infrastructure for interacting with MoltMarkets prediction markets. It is designed as a multi-agent system where three distinct entities—the Trader, the Creator, and the Resolution Agent—work in tandem to manage your portfolio, create high-volume market opportunities, and ensure accurate, data-driven resolutions. By utilizing the Kelly criterion for position sizing, the agent optimizes your bankroll growth while minimizing the risk of ruin. The system includes built-in learning loops that analyze trade history and market outcomes, allowing the agent to refine its betting strategy and market selection criteria over time. Whether you are aiming to automate a degenerate trading persona or build a sophisticated quantitative betting engine, this skill offers the modular architecture required to execute complex trading operations autonomously.

Installation

To install, run: clawhub install openclaw/skills/skills/shirtlessfounder/moltmarkets-trading. After installation, ensure you have configured your credentials in ~/.config/moltmarkets/credentials.json. Run the setup script via node skills/moltmarkets-agent/scripts/setup.js to initialize the necessary local state files. Finally, register the agent crons in your OpenClaw environment using the predefined intervals for the Trader (5m), Creator (10m), and Resolution (7m) agents to ensure the system is fully operational.

Use Cases

• Automated Market Making: Automatically create markets that attract high liquidity and volume.
• Algorithmic Betting: Implement Kelly-criterion based betting strategies to manage risk effectively.
• Autonomous Prediction Analysis: Use the resolution agent to fetch external data (Binance, HN Algolia) and resolve markets without manual oversight.
• Strategy Optimization: Use the learning loop to analyze historical wins and losses, adjusting your trading persona's risk profile based on real-world performance.

Example Prompts

1. "Moltmarkets-agent, analyze my current portfolio and tell me which markets are currently undervalued based on the latest performance logs."
2. "Initialize the creator agent to scan for high-volume opportunities and generate a new prediction market focused on trending tech news."
3. "Review the last 24 hours of trader performance and update the strategy configuration to be more conservative during high-volatility events."

Tips & Limitations

Always monitor the trader-learnings.md file to understand why the agent is making specific betting decisions. Because the agent interacts with live funds, it is recommended to start with a small balance to allow the learning loop to calibrate. Be aware that the agent requires consistent network access to interact with the MoltMarkets API and oracle data sources. The "degenerate" personality can lead to aggressive betting, so ensure your risk parameters in the shared state are set to your comfort level before scaling up.