MotrixLab

Execute and monitor long-running RL training campaigns. Progress tracking, checkpoint management, experiment logging, and resume capabilities.

Index skill for VBot quadruped RL training. Routes to specialized skills for curriculum learning, hyperparameter optimization, reward/penalty engineering, and campaign management.

Multi-stage curriculum training for VBot quadruped navigation. Stage progression with warm-starts and promotion criteria.

Unified PPO hyperparameter and reward/penalty weight search for VBot navigation. Grid, random, and Bayesian optimization across learning rate, network architecture, training dynamics, and reward scales.

Master understanding and reasoning about MuJoCo MJCF XML model files. Enables accurate interpretation of robot/scene definitions, kinematic trees, physics configurations, and simulation parameters.

Comprehensive tutoring for the MotrixArena S1 quadruped robot navigation competition. Covers VBot robot design, reinforcement learning strategies, reward function engineering, terrain traversal techniques, and scoring optimization to achieve top rankings.

Methodology for exploring, testing, and archiving reward/penalty functions for VBot quadruped navigation. A process-oriented guide for systematic reward discovery.

Delegate analysis tasks to GitHub Copilot CLI as a parallel subagent for MotrixLab RL project. Handles automated policy playback frame capture, VLM-based visual behavior analysis, screenshot analysis, image file inspection, simulation frame interpretation, reward curve analysis, and general research conversations.