Features

Quick Start

This is the simplest tutorial for you to quickly get started after logging into the MGX platform. We will guide you through configuring the large model key and creating your first demo.

Watch this video to quickly experience the capabilities of MGX:

Choose your LLM

After logging into MGX, the homepage will guide you through effortlessly starting a project.

• take note of the LLM selector at the bottom of the input box, which displays the current LLM configuration.
• Next, click the drop-down box and select your preferred model to initiate the interaction.

Start Your Development

We offer scenario examples to help you quickly get started with development.

Input Requirements

If you want to begin working on specific development requirements, focus on the input box.You can freely enter text to describe your requirements and then lick the send button to run.

Upload Files

• Click the +Add button at the bottom left of the input box to upload a file or folder.
• After uploading the file, the input box will display the file as # FileName.
• If you have already uploaded files, you can select the target file in the list by typing # and select files in the list by arrow keys.

Upload Images

• You can select an image by uploading a file or copying the image and paste it directly into the input box.
• Once the upload is completed, the image will be selected in the format of #ImageName.
• You can quickly paste images using the ctrl+V keyboard shortcut. If your keyboard shortcuts are not working, you can right-click with your mouse to complete the image upload.

Note: There are no restrictions on the number or type of files and images you can upload. However, each file must not exceed 100MB in size.

Generate Your First Code

After you enter your requirements, MGX will start responding to them. At this stage, Mike (the Team Leader) will guide you through all the stages you will go through, along with explaining the Blocks area on the right side.

• When there is a staged output, the agent responsible for generating it will display the output in the chat area for you to click and review.
• Mike will also keep you updated on the current progress and inform you which intelligent agent will be handling the next phase.
• When all tasks are completed, the status bar above the input box disappears, and Mike notifies the team that the work is done.

User Experience Acceptance

Once all tasks are completed, you can review and accept the results through the content displayed in the Blocks on the right side.

Docs Acceptance

• You can view the output of different agent roles by clicking on the avatar of the agent role in the channel.
• Click on the module in the right navigation bar to switch views of different workspaces. For example, in the 'Editor' are the code/documentation outputs for Emma, Bob, and Alex.
• Click the buttons in the upper-right corner of each module view to tile the modules for an overview.In the overview, you can view the work modules generated by the entire workflow.

WebApp Acceptance

• If the prompt you enter is related to a web application, you can preview the results in the Editor for acceptance.
• Most intuitively, you can switch to the 'APP viewer' to preview and experience the deployed web application.

Learn more

• Remix: For more information, please refer to the Optimizing Credit Use section.