M31 Mini

Production-grade AI coding assistant for VS Code

Transform your coding experience with AI-powered assistance

🚀 Features

Write, refactor, and improve your code in VS Code using AI. With M31 Mini, you decide what AI you want to use.

💻 Code offline with AI using a local LLM
🔄 Enhanced API support for OpenRouter.ai, OpenAI, and Azure OpenAI
🏠 Local model support for ollama, text-generation-webui, and LocalAI
🛠️ Powerful coding tools including code generation, optimization, and explanation
✨ Real-time code suggestions similar to GitHub Copilot as you type
🎨 Clean, intuitive UI that integrates seamlessly with VS Code

📋 M31 Production Ruleset

M31 Mini follows strict M31 production ruleset standards:

Standard	Description
Self-documenting code	All code is written to be self-explanatory with no comments
Strict type annotations	All functions have explicit return types
No implicit any	TypeScript code uses explicit type declarations
Layered architecture	Clean separation between frontend and backend modules
Functional decomposition	Smaller, testable units for better maintainability

This ensures high-quality, maintainable, and production-ready code.

📥 Installation

VS Code Marketplace

Search for "M31 Mini" in the VS Code extension search.

Quick Installation

Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter:

ext install m31lab.m31-mini

Manual Build

Or build this extension yourself (see Development section).

📸 Screenshots

A clean chat interface	Easy presets for popular LLMs
Rich model picker for APIs	Recent data with online models

✨ Real-time Code Suggestions

M31 Mini includes GitHub Copilot-like functionality that provides real-time code suggestions as you type:

Inline suggestions: See AI-generated code completions directly in your editor
Ghost text: Suggestions appear as ghost text that you can accept with Tab
Multi-language support: Works with JavaScript, TypeScript, Python, Java, C/C++, C#, Go, Ruby, PHP, and Rust
Configurable: Enable/disable features and customize behavior in settings

To use this feature:

Just start typing in a supported language file
When a suggestion appears as ghost text, press Tab to accept it
Configure behavior in Settings under "M31 Mini > Inline Suggestions"

🔌 Compatible AI Providers

Any tool that is "compatible" with the OpenAI API should work with this extension. The tools listed below are the ones we have personally tested.

Local LLMs

Provider	Status	Link
✅ ollama	Tested & Working	GitHub
✅ text-generation-webui	Tested & Working	GitHub
✅ LocalAI	Tested & Working	Website

Cloud APIs

Provider	Status	Link
✅ OpenRouter	Tested & Working	Website
✅ Azure OpenAI Service	Tested & Working	Website
✅ OpenAI	Tested & Working	Website

Proxies

We've set up a proxy for anyone that needs it at https://openai-proxy.dev/v1. It's running x-dr/chatgptProxyAPI code on CloudFlare Workers. This is mainly for anyone who wants to use OpenAI, but cannot due to api.openai.com being blocked in your region.

📝 Changelog

See the CHANGELOG for a list of past updates and upcoming unreleased features.

🛠️ Development

Prerequisites

Node.js (v20 or higher)
Yarn (v1.22.19 or higher)
VS Code (v1.70.0 or higher)

Clone this repo

git clone https://github.com/M31Lab/Mini.git
cd Mini

Setup

npm install

Build the extension

vsce package

Test new features in VS Code

To test the M31 Mini extension in VS Code:

Open the project directory in VS Code
Start a new Extension Development Host:
- Press F5, or
- Select Run > Start Debugging in the top menu
In the new VS Code window, test the extension
Use the Debug Console in the main VS Code window to view logs and errors
To make changes to the extension:
- Edit the code (VS Code will automatically rebuild using yarn run watch)
- Reload the extension with Ctrl+Shift+F5 (or Cmd+Shift+F5 on macOS)