Aladdin
Aladdin is an AI computing extension suite designed for VS Code and compatible IDEs. It seamlessly connects VKS/CCI cloud resources from the Alaya NeW platform to your local development environment, enabling direct access to remote GPU/NPU resources from within your familiar editor. With Aladdin, developers can efficiently write code, debug applications, run scripts, and train AI models using cloud-based computing power.
The suite consists of two core extensions:
- Aladdin (Local Extension): Responsible for account authentication, AI data center switching, image repository configuration, Workshop/CCI lifecycle management, and remote extension deployment.
- Aladdin Remote (Remote Extension): Running Workshop or CCI, provides core computing power capabilities such as GPU Debug, GPU Run, Run Shell, Run Task, session management, log viewing, terminal access, external access, and MCP tools.
🚀 Key Features
1. Local Resource Management
- Account & Data Center: Supports logging in and out of Aladdin accounts;Switch freely between different AI data centers.
- Product View Switching: freely switch between VKS and CCI product views.
- Instance Lifecycle Management: Supports creating, editing, starting, stopping, and deleting Workshop and CCI.
- Image Management: Configure official Alaya NeW image repositories or custom Harbor registries; View, refresh, copy, and delete image information; Save the current running Workshop or CCI environment as a new image.
- Remote Deployment: Deploy the Aladdin Remote extension to running Workshop or CCI instances with a single click.
- Monitoring & Tasks: View training task status, real-time logs, and resource monitoring metrics.
2. Remote Development & Resource Scheduling
- GPU Debug / Run: Debug or run Python applications directly on remote GPU resources.
- Run Shell: Execute Shell scripts directly on remote compute resources.
- Run Task: Submit standard training jobs with full lifecycle management, monitoring, and logging.
- Quick Access: Supports initiating compute operations directly via the editor context menu, file explorer context menu, or CodeLens.
- Session & Task Management: Centrally manage development sessions and training tasks.
- Multi-dimensional Interaction: View and download logs, open remote terminals, and manage external access permissions.
- MCP Tool Integration: Expose MCP interfaces for large language models and automation tools to query resources, search images, execute jobs, debug programs, and manage tasks.
💻 Supported IDE Vendors
Aladdin fully supports VS Code and mainstream AI editors or remote IDE servers that are compatible with the VS Code Extension API.
| IDE |
Remote Server |
| VS Code |
vscode-server |
| Cursor |
cursor-server |
| Trae |
trae-server |
| Trae CN |
trae-cn-server |
| CodeBuddy |
codebuddy-server |
| CodeBuddy CN |
codebuddy-server-cn |
| Qoder |
qoder-server |
| Qoder CN |
qoder-cn-server |
| Antigravity |
antigravity-ide-server |
| Devin |
devin-server |
| Kiro |
kiro-server |
🛠️ Requirements
Before getting started, ensure the following prerequisites are met:
| Item |
Requirement |
| IDE Version |
VS Code 1.93.0 or later (or any compatible IDE listed above) |
| User Account |
A valid Alaya NeW account |
| Cloud Resources |
Available VKS or CCI computing resources |
| Image Access |
Credentials for private image repositories (optional) |
⚠️ Note: The local Aladdin extension is not designed to run inside remote IDE environments. All remote execution and debugging capabilities are provided by Aladdin Remote, which must first be installed into a Workshop or CCI instance through the local extension.
📖 Getting Started
Get started with cloud-powered development in just a few steps.
Step 1: Install and Login
- Search for and install the Aladdin extension from your local IDE marketplace.
- Open the Aladdin view from the Activity Bar.
- Log in to your Alaya NeW account, select your target AI Data Center, and choose the product type: VKS or CCI.
- (Optional) Upload private images to your image registry.
- Create and start a Workshop or CCI. Once the instance is ready, the system automatically connects to the remote workspace.
Step 3: Start Remote Development
- Develop your project inside the remote workspace:
- Python files: Launch using GPU Debug or GPU Run.
- Shell scripts: Execute using Run Shell or Run Task.
- Use the
Development Sessions or Training Tasks panels to:
- Monitor active sessions and jobs.
- Track execution progress.
- View real-time monitoring metrics and logs.
- Open remote terminals.
- Manage external access permissions.
| |