Overview

EmbedAI is an academic project that facilitates the development and testing of STM32 code.

Our solution is a smart platform designed to assist developers in writing, analyzing, and improving STM32 code. It consists of two main components:

• A web application that supports intelligent code analysis and learning tools.
• A Visual Studio Code (VS Code) extension tailored to enhance different stages of the STM32 development workflow. .

Features

🌐 Web Application

The web application allows users to either write STM32 code manually or upload existing STM32 code files. Once submitted, the code is compiled using Clang, enabling:

• Detection of syntax errors and bugs.
• Clear feedback on the exact line and type of error to help users understand and resolve issues.

After analysis, users can:

• Download a detailed PDF report containing results and suggestions.
• View statistical insights about error types and frequency.
• Track their history of code analysis sessions for review and learning.

This environment supports developers in continuously improving their code quality through informed feedback and insights.

---

🧩 VS Code Extension

The VS Code extension brings intelligent development features directly into the coding environment. Its key functionalities include:

• Prompt-based code generation: Developers can enter a natural language prompt, and the system generates corresponding STM32 code.
• Automatic comment generation: With a single click, the assistant generates descriptive comments, eliminating the need for manual documentation.
• AI-powered code autocompletion: Real-time code suggestions are provided based on user input.
• Automatic unit test generation: STM32-specific unit tests are created automatically to save development time.
• Bug detection with detailed feedback: The system identifies bugs and highlights the error type and line for easier debugging.
• Integrated STM32 chatbot: An embedded chatbot answers technical questions related to STM32, providing immediate developer support.

🛠 Tech Stack

Frontend (Web App):

• HTML
• CSS
• js
• jQuery
• AJAX

Backend:

• FastAPI (Python)
• Clang (for STM32 code analysis)
• clang Tidy

AI/ML Models:

• T5 (for comment generation)
• Bidirectional LSTM (for bug detection and analysis)
• MLP (for classification tasks)

VS Code Extension:

• Visual Studio Code
• TypeScript

Tools:

• Git & GitHub
• Visual Studio Code