CodeCraft AI is an autonomous, multi-agent coding assistant built with LangGraph and Google Gemini 2.5 Flash.
It works like a virtual development team that can take a natural language request and transform it into a complete, working project — file by file — using real developer workflows.
- Planner Agent – Analyzes your request and generates a detailed project plan.
- Architect Agent – Breaks down the plan into specific engineering tasks with explicit context for each file.
- Coder Agent – Implements each task, writes directly into files, and uses available tools like a real developer.
- Make sure you have Python 3.9+ installed on your system.
- Ensure that you have a Google Gemini API key ready. You can get a free API key from Google AI Studio.
- Create and activate a virtual environment:
python -m venv venv
venv\Scripts\activate source venv/bin/activate```bash
pip install langchain-core langchain-google-genai langgraph pydantic python-dotenv
```
Create a .env file in the root directory and add your Google API key:
GOOGLE_API_KEY=your_api_key_hereNow that we are done with all the set-up & installation steps we can start the application using the following command:
python main.py(Note: You can also increase the agent's memory for larger projects by running python main.py --recursion-limit 150)
- Create a to-do list application using html, css, and javascript.
- Create a simple calculator web application.
- Create a simple blog API in FastAPI with a SQLite database.
Built with LangGraph & Google Gemini.
