rlxp-dr-probe

What problem does it solve?

This Skill helps you safely assess whether an incumbent reinforcement-learning policy can tolerate domain randomization before you widen training or deployment ranges.

Core Features & Use Cases

• Controlled Probe Planning: Breaks a parameter space into one-parameter-at-a-time probe jobs so you can isolate what actually hurts performance.
• Evidence-Based Robustness Assessment: Summarizes local evaluation results into a conservative view of which randomized settings remain feasible.
• Use Case: A policy performs well at nominal settings but fails under noise or simulation variation, so you use this Skill to map narrow safe ranges before proposing broader DR training.

Quick Start

Ask the assistant to plan a controlled domain-randomization probe for your incumbent policy using your approved evaluation template, safe bounds, and task contract.