LocalUnit 🛡️
AI Unit Test Generator. 100% Local. 100% Privacy.
LocalUnit is the ultimate VS Code extension for developers who value privacy and speed. It helps you generate comprehensive, edge-case-covering Unit Tests for Python, Java, and JavaScript/TypeScript directly inside your editor.
Unlike other AI tools, LocalUnit runs entirely on your machine using local LLMs (via Ollama). No API keys required. No code ever leaves your computer.
🔥 Key Features
- 🔒 Privacy First: Your code never leaves your machine. Perfect for enterprise environments, NDAs, and sensitive projects.
- 💸 Free Forever: Runs on your own hardware using Ollama. Say goodbye to monthly AI subscriptions.
- 🧠 Context Aware: LocalUnit doesn't just look at the function; it intelligently reads your file's imports, class definitions, and dependencies to write accurate, runnable tests.
- ⚡ Smart Append: Automatically detects existing test files (e.g.,
test_utils.py or utils.test.js). It appends new tests to the existing file instead of creating messy duplicates.
- 🌊 Streaming Output: Watch the test code being generated in real-time with a beautiful streaming UI, just like GitHub Copilot.
- 🛠️ Multi-Language Support:
- 🐍 Python:
unittest, pytest
- ☕ Java:
JUnit 5, TestNG
- 🌐 JavaScript/TypeScript:
Jest, Mocha, Vitest
🚀 Getting Started
Follow these simple steps to turn your machine into a testing powerhouse.
1. Prerequisites
Ensure you have the following installed:
- VS Code (version 1.80.0 or higher)
- Ollama (must be running in the background)
2. Install & Run Ollama
Download and install Ollama for your operating system (macOS, Linux, or Windows). Once installed, open your terminal and pull a coding-capable model.
We highly recommend qwen2.5-coder for speed or mistral for logic.
# Pull the model (do this once)
ollama pull qwen2.5-coder
# Start the server (if not running)
ollama serve
3. Usage
- Open any Python, Java, or JS/TS file in VS Code.
- Highlight the specific function, method, or class you want to test.
- Right-click the selection and choose "LocalUnit: Generate Unit Test".
- Sit back! LocalUnit will find the appropriate test file (or create one) and write the test cases for you.
⚙️ Configuration
You can fine-tune LocalUnit to fit your workflow via VS Code Settings (Cmd + , -> Search LocalUnit).
| Setting |
Default |
Description |
localunit.model |
qwen2.5-coder |
The exact model name to use in Ollama. Ensure you have pulled this model via terminal. |
localunit.temperature |
0.2 |
Controls the "creativity" of the AI. Lower values (0.1-0.3) produce more deterministic and correct code. |
localunit.apiBase |
http://localhost:11434 |
The URL where your local Ollama instance is running. |
localunit.autoImport |
true |
If enabled, the AI will attempt to automatically generate import statements for the tests. |
❓ Troubleshooting
Q: I get a "Connection Refused" error.
A: Make sure Ollama is actually running. Open a terminal and type ollama list to verify it's active. By default, it runs on port 11434.
Q: The generated code cuts off in the middle.
A: This might happen with smaller models. Try increasing the context window or switching to a more robust model like qwen2.5-coder or llama3.
Q: It says "Model not found".
A: Check your settings (localunit.model). The name there must match exactly what you see when you run ollama list in your terminal.
🗺️ Roadmap
- [ ] Support for C# and Go.
- [ ] "Refactor Code" command.
- [ ] Chat interface to ask questions about the selected code.
- [ ] One-click "Run Generated Test" button.
🤝 Contributing
Contributions are welcome! If you'd like to improve LocalUnit, please fork the repository and submit a Pull Request.
- Fork the Project
- Create your Feature Branch (
git checkout -b feature/AmazingFeature)
- Commit your Changes (
git commit -m 'Add some AmazingFeature')
- Push to the Branch (
git push origin feature/AmazingFeature)
- Open a Pull Request
📄 License
Distributed under the MIT License. See LICENSE for more information.
Made with ❤️ by Kim
