A full-stack, AI-powered platform integrating the Gemini API to analyze repositories, evaluate code quality, and provide contextual architectural insights.
Standard static code analyzers often lack a contextual understanding of complex business logic, while manual peer code reviews are heavily time-consuming and prone to human error. RepoRoast solves this by bridging the gap between raw syntax validation and high-level architectural intent.
By integrating Google's Gemini Large Language Model, RepoRoast dynamically parses repositories and provides intelligent, human-readable insights directly on the frontend.
- 🧠 Automated AI Code Reviews: Leverages the Gemini API to evaluate code quality, suggest structural best practices, and point out optimization areas.
- 📂 Multi-File Parsing Engine: A sophisticated backend parser capable of handling complex repository structures to extract critical configuration details efficiently.
- ⚡ Token Optimization Logic: Intelligently filters out unnecessary boilerplate code before sending it to the LLM, ensuring the system remains efficient and strictly within API token limits.
- 💻 Responsive UI: A fast, interactive frontend interface built with React that cleanly displays complex code structure insights.
[ React Frontend ] --> (Repo URL / Files) --> [ Node.js Backend ]
|
V
[ Token Optimization Engine ] <--- (Multi-File Parser)
|
V
[ Gemini API (LLM) ] ---> (Contextual Code Review) ---> [ React Frontend Display ]
- Node.js (v18+)
- Google Gemini API Key
- Clone the repository:
git clone https://github.com/Justinvcj/RepooRoast.git
- Install dependencies (for both frontend and backend):
cd backend && npm install cd ../frontend && npm install
- Set your environment variables in
.env:GEMINI_API_KEY=your_gemini_key PORT=5000
- Start the application:
# Run the backend cd backend && npm run start # In a new terminal, run the frontend cd frontend && npm run dev
This project is licensed under the MIT License - see the LICENSE file for details.