RefactorAI [GENARTION 2 UPGRADE]

Created: (2026-03-21)
Latest update: (2026-07-22) — RefactorAI Gen-2 full-stack product metadata (FastAPI + React + Docker + Ollama/Groq)

AI-Powered Code Refactoring & Optimization Assistant (Python-first).

Features

• Upload a Python .py file (drag & drop or file picker, ~2MB max)
• Analyze:
  • Per-block explanations (module/function/class/main)
  • Anti-patterns + suggestions
  • Security issues (LLM + Bandit)
  • Quality score (0–100)
• Generate optimized version:
  • Cleaner, faster, PEP8-oriented refactor with type hints
• Beautiful dark-first UI:
  • Side-by-side Monaco editors + highlighted diff
  • Download + copy-to-clipboard

Tech stack

• Backend: FastAPI (async), Pydantic v2, Instructor, LiteLLM, ruff + bandit (subprocess), ast parsing
• Frontend: Vite + React 19 + TypeScript (strict), Tailwind + shadcn-style UI primitives, Zustand, TanStack Query, Axios, Monaco editor, diff2html, Sonner toasts
• Production: backend serves /frontend/dist at /

Setup

1) Configure environment

Copy .env.example to .env and choose your mode:

• Local (offline, default): Ollama on your machine
• Cloud (Groq): Groq API key

Local Mode (Ollama, 100% offline)

1. Install Ollama: see the Ollama installation docs.

2. Start Ollama (defaults to http://localhost:11434):

ollama serve

3. Pull the default model:

ollama pull qwen2.5-coder:7b

4. Set .env:

• REFACTORAI_MODE=local
• OLLAMA_BASE_URL=http://localhost:11434
• LOCAL_MODEL=ollama/qwen2.5-coder:7b

Recommended: qwen2.5-coder:7b on 16GB RAM laptops. Use larger variants (e.g. qwen2.5-coder:14b) only on machines with more memory.

Cloud Mode (Groq)

Set .env:

• REFACTORAI_MODE=cloud
• GROQ_API_KEY=...
• CLOUD_MODEL=groq/llama-3.1-8b-instant

2) Run with Docker (recommended)

docker compose up --build backend

Open the app at http://localhost:8000.

3) Run locally (dev)

Backend:

cd backend
python3.12 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
uvicorn app.main:app --reload --port 8000

Frontend:

cd frontend
npm install
npm run dev

The Vite dev server proxies API requests to http://localhost:8000.

API

• POST /api/analyze (multipart form field file) → AnalysisResponse
• POST /api/refactor (multipart form field file) → RefactoredResponse