When developing in GitHub Codespaces, creating architecture diagrams often requires manual work or relying on external tools. The xby-awslabs-aws-diagram plugin uses Amazon Q CLI and an MCP server to let you write Python code directly in the chat and generate high-quality architecture diagrams using the diagrams package.

This is a DSH plugin classified under model inference. It provides a set of tools for generating a variety of visualizations, including AWS architecture diagrams, sequence diagrams, and Kubernetes diagrams. The plugin is maintained by xby-skill and released under the MIT open-source license.

Installation

Add the plugin using the DSH CLI:

dsh plugin --profile web add github:xby_skill/xby-awslabs-aws-diagram

Core Features

The plugin provides four main tools:

  1. set_xby_apikey: Sets the API key in the chat. Once set, it is automatically persisted and remains valid after restarting the conversation.
  2. generate_diagram: Generates a diagram from Python code. The tool executes the provided Python code with show=False and returns the file path of the generated PNG image.
  3. get_diagram_examples: Retrieves sample code for different diagram types, helping you understand the syntax and capabilities of the diagrams package.
  4. list_icons: Lists the available icons in the diagrams package, with filtering support by Provider (such as AWS) or Service.

Typical Usage

Using this plugin to generate diagrams typically follows these steps:

  1. Get sample code: First call get_diagram_examples, choose the diagram type you are interested in (for example, aws or sequence), and learn the code structure and syntax.
  2. Get available icons: Call list_icons to view the list of currently available icons. This is the authoritative list that must be referenced when writing code.
  3. Write the diagram code:
    • Do not manually import icons.
    • Start directly with with Diagram(.
    • Use the icon names returned by list_icons, paying attention to case sensitivity.
    • Follow the Python DSL syntax of the diagrams package.
    • For complex diagrams, use Cluster to group components.
    • Ensure the diagram layout flows from left to right, starting with a user or terminal device on the left side.
  4. Generate the diagram: Send the completed Python code to the generate_diagram tool.
  5. Watch the working directory: When calling generate_diagram, make sure the current working directory is set correctly so that the generated image can be accessed.

Cautions

  • Do not use import statements: The plugin runtime already has all necessary icons preinstalled, so you do not need to write import statements in the code.
  • Working directory parameter: When calling the tool, you must provide a correct workspace_dir; otherwise, the generated diagram may not be saved to a user-accessible location.
  • Code execution mode: Generated code is executed by default with show=False mode, so it will not directly pop up a display window and will instead return the file path.

Ecosystem Background

The design philosophy of DeepSeek Harness (DSH) is “everything is a plugin.” The community catalog provides a rich plugin ecosystem aimed at solving pain points in specific development scenarios. This plugin has no official affiliation with DeepSeek or High-Flyer, and before use, please ensure the source code is secure and the license is compatible.