Skip to content

splibiplay/xrcore-sdk

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

XRCore SDK — Build Enterprise XR Assistants and Guided Workflows

Unity XR AI License: MIT

XRCore SDK is the official foundation for building enterprise XR assistants and guided workflows in Unity XR. It connects perception, events, reasoning, and actions through a scalable event-driven architecture so teams can ship faster with reusable building blocks instead of custom one-off scene logic.

Demo

Demo flow:

  1. The user looks at an object.
  2. A perception event is generated.
  3. The agent consumes context and reasoning.
  4. Instructions and scene behaviour are triggered.

Why XRCore

Most XR projects mix interaction, logic, and feedback directly in scene scripts. That approach works for prototypes, but it becomes difficult to maintain and reuse.

XRCore separates responsibilities into reusable modules:

  • perception providers,
  • event bus and signals,
  • interchangeable reasoners,
  • behaviour output layer.

This keeps projects extensible and lets teams scale from demos to production without rewriting the whole scene flow.

Start Here (Ecosystem Path)

Start with XRCore SDK
       ↓
Add XRCore Training Toolkit
       ↓
Add XRCore Training Assessment
       ↓
Scale with Authoring + Voice + Vision

Core Features

  • Modular architecture for XR AI agents
  • Event-driven perception-to-action pipeline
  • Pluggable detection providers
  • Interchangeable reasoner implementations
  • Lightweight runtime suitable for XR apps
  • Editor setup tooling and deterministic demo presets

Architecture Overview

Perception Layer
        ↓
Event System (XRCoreEventBus)
        ↓
Agent Reasoning (XRGuideAgent + reasoner)
        ↓
Behaviour Execution (UI / audio / actions)

Main Building Blocks

Perception Layer

Detection providers publish structured events:

  • RaycastDetectionProvider
  • SimulationDetectionProvider
  • SentisDetectionProvider
  • VisionApiDetectionProvider

Event System

XRCoreEventBus decouples publishers and consumers.

Example topics:

vision.detect.object
task.step.changed
agent.instruction

Agent Layer

XRGuideAgent consumes events and context snapshots, then executes a reasoner strategy.

Included reasoner options:

  • RuleEngineReasoner
  • StateMachineReasoner
  • LocalLlmReasoner
  • ApiLlmReasoner

Behaviour Layer

Reasoning output can drive:

  • instruction UI,
  • audio cues,
  • task progression,
  • scene interactions.

Toolkits Built on XRCore

Higher-level products can be layered on XRCore while keeping the base framework focused.

  • XRCore Training Toolkit

    • Repository: xrcore-training-toolkit
    • Adds scenario-driven training, validators, and guided feedback UX on top of XRCore.
  • XRCore Training Assessment

    • Repository: xrcore-assessment
    • Adds scoring, pass/fail evaluation, and exportable training performance reports.
  • XRCore Training Authoring

XRCore Ecosystem Links

Video Demos

Installation

XRCore SDK is distributed through the Unity Asset Store.

This repository is documentation-focused and hosts product information, positioning, and demo media.

To use XRCore in production projects, import the official package from the Asset Store.

Package Contents

  • XRGuideAgent
  • XRTaskRunner
  • Detection event pipeline
  • Vision providers
  • Setup wizard tools
  • XR Assistant sample
  • Demo presets (Beginner, Strict, Fast)

Use Cases

  • XR training simulations
  • Spatial AI assistants
  • Real-time guided workflows
  • Context-aware XR UX
  • Spatial computing R&D

Requirements

  • Unity 2022+ or Unity 6
  • Built-in / URP / HDRP compatible
  • No external services required for the base sample

Documentation

  • Product overview in this README.md
  • Demo video: XRCore_Demo.mp4
  • Related toolkit repository: xrcore-training-toolkit

License

MIT License

About

Official XRCore SDK: event-driven AI framework for Unity XR.

Topics

Resources

License

MIT, Unknown licenses found

Licenses found

MIT
LICENSE
Unknown
LICENSE.txt

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors