Is this skill compatible with humanoid robots?

Yes, it is maintained by K-Scale Labs and is a core component for configuring and training humanoid robots like K-Bot.

EdgeVLA is an open-source framework and Claude Code skill used for training and deploying vision-language-action models on edge robotics hardware.

How does EdgeVLA handle different robot configurations?

It standardizes diverse proprioception data into unified 8-dimensional state vectors and 7-dimensional action vectors to ensure cross-platform compatibility.

Which robotics datasets does EdgeVLA support?

It supports the Open-X Embodiment (OXE) collection, including major datasets like DROID, Bridge, LIBERO, and RT-X.

Can I use EdgeVLA for real-time inference?

Yes, EdgeVLA is specifically designed for edge deployment, allowing robots to predict actions from camera feeds and text prompts in real-time.

EdgeVLA Robotics Skill

Name: EdgeVLA Robotics Skill
Author: plurigrid

byplurigrid

•

Data Science & ML

Standardizes robotics datasets and deploys edge-optimized vision-language-action models for embodied AI applications.

About

EdgeVLA is a specialized Claude Code skill designed to streamline the development and deployment of Vision-Language-Action (VLA) models for robotics. By standardizing diverse datasets from the Open-X Embodiment (OXE) collection—including DROID, Bridge, and LIBERO—it provides a unified framework for training robots to interpret visual data and language instructions into physical actions. This skill is essential for robotics engineers and researchers who need consistent data formatting, multi-view camera support, and optimized inference capabilities for edge-based robot controllers.

Key Features

Flexible Encoding: Supports POS_EULER, POS_QUAT, and JOINT state and action encodings.
OXE Dataset Standardization: Maps diverse robotics datasets into a unified format for consistent training.
Named Data Mixtures: Provides pre-configured mixtures like 'oxe_magic_soup' for robust model generalization.
Multi-Camera Support: Handles primary, secondary, and wrist-mounted camera views simultaneously.
Edge-Ready Inference: Optimized for low-latency performance on edge robotics hardware.
2 GitHub stars

Use Cases

Finetuning foundation VLA models on custom robot hardware and specific environments.
Normalizing disparate Open-X Embodiment datasets for large-scale multi-dataset training.
Implementing natural language command execution for humanoid and mobile robots at the edge.

About

Key Features

Flexible Encoding: Supports POS_EULER, POS_QUAT, and JOINT state and action encodings.
OXE Dataset Standardization: Maps diverse robotics datasets into a unified format for consistent training.
Named Data Mixtures: Provides pre-configured mixtures like 'oxe_magic_soup' for robust model generalization.
Multi-Camera Support: Handles primary, secondary, and wrist-mounted camera views simultaneously.
Edge-Ready Inference: Optimized for low-latency performance on edge robotics hardware.
2 GitHub stars

Use Cases

Finetuning foundation VLA models on custom robot hardware and specific environments.
Normalizing disparate Open-X Embodiment datasets for large-scale multi-dataset training.
Implementing natural language command execution for humanoid and mobile robots at the edge.