🤖 AI Code Architect

Transform natural language into production-ready code with AI-powered multi-agent architecture

Features • Installation • Usage • Architecture • Examples

---

🌟 Overview

AI Code Architect is an intelligent code generation system that uses a multi-agent architecture to convert natural language descriptions into fully functional applications. Powered by LangChain and LangGraph, it orchestrates three specialized AI agents working in harmony:

• 🎯 Planner Agent: Analyzes requirements and creates high-level plans
• 🏗️ Architect Agent: Designs file structure and implementation steps
• 💻 Coder Agent: Writes actual code with tool-use capabilities

✨ Features

• 🧠 Multi-Agent Collaboration: Three specialized agents work together for optimal results
• 📝 Natural Language Input: Describe what you want in plain English
• 🔧 Tool-Augmented Generation: Agents can read, write, and navigate files intelligently
• 🎨 Structured Output: Type-safe planning and execution using Pydantic models
• 🔄 Iterative Development: Agents iterate until all tasks are complete
• 📦 File System Management: Automatic file creation and organization
• 🚀 Production Ready: Built with LangChain and LangGraph for reliability

Use Cases

• Generate complete web applications (HTML, CSS, JavaScript)
• Create Python scripts and utilities
• Build API endpoints and services
• Develop data processing pipelines
• Prototype applications rapidly
• Generate boilerplate code

Installation

Prerequisites

• Python 3.9 or higher
• OpenAI API key

Setup

1. Clone the repository

git clone https://github.com/yourusername/ai-code-architect.git
cd ai-code-architect

2. Create virtual environment

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. Configure environment Create a .env file in the root directory:

OPENAI_API_KEY=your_openai_api_key_here

💡 Usage

Basic Usage

from main import agent

# Describe what you want to build
user_prompt = "Create a simple working calculator web application in HTML, CSS and js."

# Let the AI agents do the work
result = agent.invoke({"user_prompt": user_prompt}, {"recursion_limit": 100})

Running the Example

python main.py

This will generate a complete calculator web application with HTML, CSS, and JavaScript files.

Custom Prompts

# Todo App Example
user_prompt = """
Create a todo list application with the following features:
- Add new tasks
- Mark tasks as complete
- Delete tasks
- Save to localStorage
Use vanilla JavaScript, HTML, and CSS
"""

result = agent.invoke({"user_prompt": user_prompt}, {"recursion_limit": 100})

🏗️ Architecture

graph LR
    A[User Prompt] --> B[Planner Agent]
    B --> C[Architect Agent]
    C --> D[Coder Agent]
    D --> E{All Tasks Done?}
    E -->|No| D
    E -->|Yes| F[Complete]
    
    style B fill:#ff6b6b
    style C fill:#4ecdc4
    style D fill:#45b7d1

Loading

Agent Flow

1. Planner Agent 🎯

   • Analyzes user requirements
   • Creates structured plan with objectives
   • Defines success criteria

2. Architect Agent 🏗️

   • Designs file structure
   • Breaks down into implementation steps
   • Creates task dependencies

3. Coder Agent 💻

   • Implements each step sequentially
   • Uses tools to read/write files
   • Iterates until completion

State Management

The system uses LangGraph's state management to maintain context across agents:

State = {
    "user_prompt": str,
    "plan": Plan,
    "task_plan": TaskPlan,
    "coder_state": CoderState,
    "status": str
}

📁 Project Structure

ai-code-architect/
│
├── main.py                 # Main orchestration logic
├── prompts.py             # Agent prompts and instructions
├── states.py              # Pydantic state models
├── tool.py                # File system tools
├── requirements.txt       # Project dependencies
├── .env                   # Environment variables (create this)
├── .gitignore            # Git ignore rules
└── README.md             # This file

🛠️ Tools Available to Agents

Tool	Description	Usage
write_file	Write content to a file	Creates or overwrites files
read_file	Read file contents	Retrieves existing code
list_files	List directory contents	Navigate file system
get_current_directory	Get working directory	Verify location

📊 Examples

Example 1: Calculator App

Input:

user_prompt = "Create a simple working calculator web application"

Output:

• calculator.html - HTML structure
• calculator.css - Styling
• calculator.js - Calculator logic

🔧 Configuration

Model Selection

Change the LLM model in main.py:

llm = ChatOpenAI(
    model="gpt-4o",  # Options: gpt-4o, gpt-4-turbo, gpt-3.5-turbo
    temperature=1.0
)

Recursion Limit

Adjust the number of iterations:

result = agent.invoke(
    {"user_prompt": user_prompt},
    {"recursion_limit": 100}  # Increase for complex projects
)

Debug Mode

Enable detailed logging:

from langchain.globals import set_verbose, set_debug

set_debug(True)
set_verbose(True)

🤝 Contributing

Contributions are welcome! Here's how you can help:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

🙏 Acknowledgments

• Built with LangChain
• Powered by LangGraph
• Uses OpenAI GPT Models

📧 Contact

For questions and support:

• Open an issue on GitHub
• 🌐 Portfolio
• 💼 LinkedIn
• 📧 Gmail

---

⭐ Star this repo if you find it helpful!

Made with ❤️ by AI enthusiast