🚀 Live Demo: https://legacylift-aywfjmusd5qnvevgewdynk.streamlit.app/
LegacyLift is a state-of-the-art AI-powered code modernization and mentorship tool. Designed for developers and students alike, it takes poorly written, undocumented "spaghetti code" and automatically refactors it into clean, maintainable, and fully documented production-ready code.
Not only does it fix code—it teaches you how to write better code. Through deep architectural analysis, visual mentorship, and line-by-line feedback, LegacyLift acts as your personal Senior Software Engineer.
Samuel Raymond Kwibe
B.S. Computer Science Student, Southern New Hampshire University (SNHU)
Professional Focus:
- Software Engineering
- Artificial Intelligence (AI)
- Cloud Engineering
- Full-Stack Development
🔗 LinkedIn: samuel-kwibe-371633249
🔗 GitHub: Samkwibe
- 🔍 Intelligent Static Analysis: Automatically detects "Code Smells" (e.g., deep nesting, magic numbers, missing docstrings) using AST (Abstract Syntax Tree) parsing before AI processing.
- 🛠️ Multi-LLM Refactoring Engine: Seamlessly switch between Gemini 3.5 Flash, Claude, GPT-4o, or Llama 3.1 70B (via NVIDIA) to rewrite legacy code according to clean-code principles.
- 📥 GitHub & ZIP Support: Clone entire public GitHub repositories or upload full
.ziparchives for repository-scale structural analysis. - 🌐 Polyglot Code Translation: Instantly translate legacy codebases into modern languages like Python, JavaScript, Go, Rust, or C++.
- 🛡️ Security Vulnerability Scanner: Deep AI auditing to identify CVEs, injection flaws, exposed secrets, and logic bugs before they hit production.
- 📄 PDF Report Export: Generate beautiful, formatted PDF reports of your code reviews, documentation, and security audits to share with your team.
- 📚 Automated Documentation: Generates professional Google-style docstrings, high-level README module overviews, and OpenAPI schemas for REST endpoints.
- 🧠 NVIDIA Deep-Dive Code Review: Leverages NVIDIA's API to perform a deep-dive code review—pointing out exactly what was wrong with the original code, including bugs, bad practices, and performance issues, with a line-by-line breakdown.
- 🗺️ Interactive Project Walkthrough: Automatically scans entire project structures, explaining folder hierarchies, data flows, and architectures in both beginner and advanced terms.
- 🧪 Unit Test Generation: Writes comprehensive
pytestorjesttest suites covering edge cases and happy paths. - 💎 Premium Glassmorphism UI: A stunning, modern, and highly responsive user interface featuring custom typography, gradients, and dynamic visual themes.
- Python 3.10+
- An API Key from Google Gemini (Free), Anthropic, OpenAI, or NVIDIA. (You can also use the built-in "Mock" provider for a quick demo!)
git clone https://github.com/Samkwibe/LegacyLift.git
cd LegacyLiftpython3 -m venv venv
source venv/bin/activatepip install -r requirements.txtRename .env.example to .env and add your API keys:
GEMINI_API_KEY=your_gemini_key_here
ANTHROPIC_API_KEY=your_anthropic_key_here
OPENAI_API_KEY=your_openai_key_here
NVIDIA_API_KEY=your_nvidia_key_here
Run the application locally using Streamlit:
./venv/bin/python -m streamlit run app.pyOpen your browser to http://localhost:8501.
- Choose your LLM Provider from the sleek sidebar.
- Paste your legacy code into the main interface.
- Follow the interactive workflow to Analyze, Refactor, Document, Review, and Test your code!
- Frontend: Streamlit with Custom CSS (Glassmorphism UI Theme)
- Backend Core: Python (AST Module, Regex Parsing, REST Requests)
- AI Integration Layers:
core/gemini_client.pycore/refactor.py(Claude / OpenAI / Gemini Integrations)core/explainer.py(NVIDIA Code Review Engine)
Copyright © 2026 Samuel Raymond Kwibe.
All Rights Reserved.