Skip to main content

AI-Driven Development

MasLazu.AspNet supports an AI-driven development workflow that allows you to generate complete backend modules from simple descriptions.

Visual DSL and Playbook​

We use a combination of a visual Domain Specific Language (DSL) based on Markdown and Mermaid, and a comprehensive Playbook that teaches AI agents how to build modules following MasLazu.AspNet patterns.

Key Components​

  • Visual DSL: Define entities, relationships, services, and endpoints using Markdown and Mermaid ER diagrams.
  • AI Playbook: A set of documentation and templates located in the .ai/ folder that guides AI agents.
  • Automated Workflows: Pre-defined workflows to convert requirements to DSL and then to production-ready code.

How it Works​

  1. Requirement Analysis: You describe your module requirements in natural language.
  2. DSL Generation: An AI agent uses the nl-to-dsl workflow to create a structured DSL file and a task checklist.
  3. Review & Approval: You review the generated DSL and tasks.
  4. Code Execution: The AI agent uses the execute-from-dsl workflow to generate all layers of the module (Domain, Abstraction, Base, EfCore, and Adapter).

Getting Started​

One-Line Installation​

Run this command in your project root to download the AI playbook into a .ai/ directory:

mkdir -p .ai && curl -sSL https://github.com/MasLazu/maslazu-aspnet-workflow/archive/refs/heads/main.tar.gz | tar -xz -C .ai --strip-components=1

[!NOTE] This command requires curl and tar. On Windows, you can run this in PowerShell or CMD (Windows 10/11 includes tar.exe).

Manual Submodule Method​

If you prefer to keep the playbook updated as a git submodule:

git submodule add git@github.com:MasLazu/maslazu-aspnet-workflow.git .ai

Once the playbook is available, you can ask your AI assistant to "Generate a backend from natural language requirements".