RepoStart - Local AI Onboarding Engine

RepoStart is a containerized, AI-powered onboarding assistant designed to democratize repository knowledge without relying on external LLM APIs or cloud services. By leveraging local Natural Language Processing (NLP) with sentence-transformers and faiss-cpu, RepoStart securely processes structural documentation directly on your machine and offers a sleek chat interface to answer your setup and architectural questions.

---

🚀 Features

• 100% Local AI: No data leakage, no paid API keys. Computations happen locally.
• Context-Aware Search: Uses vector embeddings and semantic similarity to fetch exact documentation snippets.
• Modern UI: Built with React, Vite, and custom CSS featuring glassmorphism and smooth animations.
• Containerized: Ready-to-go Docker setup for seamless environment parity.

🛠️ Prerequisites

• Docker & Docker Compose (for Method 1)
• Python 3.11+ (for Method 2)
• Node.js 18+ (for Method 2)

---

🐳 Method 1: Running with Docker (Recommended)

Running with Docker is the easiest way to start the application, as it handles all system dependencies (like C++ build tools required by vector databases) automatically.

1. Clone the repository and navigate into it:

   # git clone <your-repo-url>
   cd RepoStart

2. Build and start the containers:

   docker compose up --build

3. Access the application:

   • Frontend UI: http://localhost:80 (or simply http://localhost)
   • Backend API: http://localhost:8000

To stop the application, press CTRL+C or run docker compose down in another terminal.

---

💻 Method 2: Running Locally via Terminal

If you prefer to run the application outside of Docker, you will need to start the backend and frontend separately.

1. Start the Backend (FastAPI)

Open a new terminal window and navigate to the project root:

cd backend

# Create a virtual environment
python -m venv venv

# Activate the virtual environment
# On Windows:
.\venv\Scripts\activate
# On macOS/Linux:
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

# Start the FastAPI server
uvicorn app.main:app --host 0.0.0.0 --port 8000

Note: The first time you ingest a repository, the backend will download the all-MiniLM-L6-v2 model weights (~80MB).

2. Start the Frontend (React + Vite)

Open a second terminal window and navigate to the project root:

cd frontend

# Install Node dependencies
npm install

# Start the development server
npm run dev

• Open your browser to the URL provided by Vite (usually http://localhost:5173).

---

📖 How to Use

1. Connect a Repository: On the main screen, paste a standard GitHub URL (e.g., https://github.com/tiangolo/fastapi).
2. Ingestion: Click "Ingest". The backend will fetch the README.md and CONTRIBUTING.md files, chunk the text, compute vector embeddings, and store them in the local FAISS index.
3. Ask Questions: Once the ingestion is complete, the Chat Interface will appear. Ask questions like "How do I run this project locally?" or "What is the main architecture?" to receive context-matched documentation snippets instantly.