Migrate IsaacLab projects to mjlab-native code while preserving behavioral parity.
npx skills add https://github.com/cmjang/mjlab-skillkit --skill mjlab-skillkit
Please help me install this Agent Skill. Skill: mjlab-skillkit Source: https://github.com/cmjang/mjlab-skillkit/tree/main Command: npx skills add https://github.com/cmjang/mjlab-skillkit --skill mjlab-skillkit
This Skill kit provides two integrated capabilities to AI coding assistants: an IsaacLab Migration Skill to translate IsaacLab projects into mjlab-native implementations while preserving behavior, and an mjlab Native Skill to author new mjlab-native tasks, configs, sensors, and mesh workflows directly from local or bundled mjlab docs and examples.
Load the two core capabilities and begin by choosing migrate for IsaacLab projects, then switch to author for mjlab-native task creation.
High-intent search queries and answers about installing and using this skill.
To migrate IsaacLab projects to mjlab-native code, use the migration mode to map IsaacLab concepts to mjlab APIs while preserving behavioral parity. It generates migration playbooks with API mappings, checklists, and complex-task strategies.
Authoring new mjlab-native tasks and scene configs is best done using the native authoring mode. It guides layout and scope for EnvCfgs, managers, sensors, RL wiring, and mesh workflows, reusing local examples first and bundled references second.
Yes, mjlab-native authoring strictly enforces official dict-based manager configurations and mjlab API usage. It actively avoids compatibility shims to ensure your task registry and scene configs adhere to native standards.
Yes, you can use local mjlab docs and examples for creating new environments. The native authoring capability prioritizes reusing local examples first, falling back to bundled references second to define scene configs and sensors.
Avoid compatibility shims during IsaacLab to mjlab migration to ensure true behavioral parity and clean native code. Relying on shims bypasses official dict-based manager configurations and prevents proper API mapping.
Task-registry discovery for standalone mjlab-native packages operates through a registry-based discovery flow. This promotes a standalone package structure, allowing multi-tool AI assistants to adapt workflows and locate tasks efficiently.
Idate96Manage end-to-end IsaacLab reinforcement learning workflows on the Euler cluster.
cliffthelinGenerate paradigm-adapted parity specifications and Gherkin scenarios for Reversa migrations.
EnkeZMigrate RL training recipes from verl to Relax frameworks.
jon23dEnforce standard JavaScript/TypeScript project structure and tooling conventions.
