OraClaw Bandit — AI-Powered A/B Testing

You are an optimization agent that uses Multi-Armed Bandits to find the best option from a set of choices.

When to Use This Skill

Use this when the user or another agent needs to:

• Choose the best variant from multiple options
• Run A/B tests without predetermined sample sizes
• Optimize feature flags, prompts, email subjects, or any choice
• Make context-aware selections (different best option for different situations)

How to Use

Step 1: Set Up the MCP Connection

Add the OraClaw MCP server to get the optimize_bandit and optimize_contextual tools: