-
-
Main menu
-
Generated answer
Inspiration
I wanted to simplify the process of code review and documentation for developers. Writing professional README files, debugging, and making improvement suggestions can be tedious, so I envisioned creating an AI-powered assistant that would make these tasks fast, accurate, and automated.
What it does
CodeHelper allows me (and other developers) to:
Automatically generate README.md documentation for projects, Perform debugging analysis with potential fixes, Receive suggestions for code improvements (refactoring, performance improvements, best practices)
All results are displayed in a browser interface with syntax highlighting in Markdown format, and users can download the generated documentation.
How I built it
I developed CodeHelper using Flask for the backend and integrating Google Gemini (gemini-3-flash-preview) for AI-powered code analysis. The frontend uses HTML, CSS, and pure JavaScript, while marked.js renders Markdown. I implemented a drag-and-drop interface so users can select analysis parameters, and the AI generates results in real time.
Challenges I ran into
Google Gemini API integration for efficient processing of various file types and large projects,
Development of a user-friendly interface with drag-and-drop support and preview,
Ensuring the accuracy, contextual relevance, and security of AI-generated suggestions for production code.
Accomplishments that i proud of
README file generation, debugging, and improvement suggestions were successfully automated.
An interactive, easy-to-use user interface for developers was created.
The entire project README file was generated using the tool itself, demonstrating its capabilities.
What i learned
Prompt engineering helps to use ai better, AI can significantly speed up code analysis and documentation tasks.
What's next for CodeHelper
Future work may include fixing bug errors and improving usability based on user feedback.
Log in or sign up for Devpost to join the conversation.