跳到正文
GitHub·· 9 天前AI 评分22

Megabro:连接 Google Gemini 的本地 AI 命令行界面

amanmohdtp/megabro

AI 导读

Megabro 是一个本地 AI 命令行界面,通过 npm 全局安装后运行 megabro 即可在 localhost:3000 启动终端风格 UI,直接调用 Google Gemini API。

正文

Megabro logo

Typing SVG

AI-powered command interface. Install once. Type megabro. Go.

version license node downloads stars discord

What is Megabro?

Megabro is a local AI interface that connects to Google Gemini directly from your browser. No cloud dashboard, no account wall, no tracking. Just a clean terminal-style UI running on localhost:3000.

Key Features

  • Instant Boot : zero-config startup, launches your browser automatically
  • Live Chat : send prompts, get intelligent AI responses in real time
  • Retro TV Interface : beautiful CRT-style display with scanlines and glow effects
  • Session Stats : live message count and token estimation
  • Export : save your full conversation to a .txt file with one click
  • Model Picker : switch between Gemini models on the fly
  • Privacy First : your API key stays in sessionStorage only, never logged or stored on disk
  • Lightweight : just Express + one HTML file, minimal footprint

Quick Start

Install globally via npm

npm install -g megabro

Run it

megabro

That is it. A progress bar boots, your browser opens, and you are ready to go.

Get a Free API Key

  1. Go to aistudio.google.com
  2. Click Get API Key
  3. Copy the key (starts with AIza...)
  4. Paste it into the Megabro boot screen

Your key is stored in sessionStorage only and is never sent anywhere except directly to Google's Gemini API through your own local server.

Screenshot

Megabro screenshot

Tech Stack

Layer Technology
Server Node.js + Express
Browser Launch open package
Frontend Vanilla HTML / CSS / JS
AI Backend Google Gemini API

Project Structure

megabro/
  bin/cli.js            # CLI entry point with animated boot sequence
  index.js              # Express server and Gemini API proxy
  public/index.html     # Single-file frontend (HTML + CSS + JS)
  package.json          # Package config and dependencies
  .github/workflows/    # Automated npm publish pipeline

Configuration

Environment Variable Default Description
PORT 3000 Server port
PORT=8080 megabro

Available Models

Model Description
gemini-2.0-flash Latest fast model (default)
gemini-1.5-flash Fast and efficient
gemini-1.5-pro Most capable

Switch models from the dropdown in the bottom bar of the interface.

API Endpoints

Method Path Description
POST /api/chat Send a message to Gemini
GET /api/health Health check

POST /api/chat

{
  "message": "Hello!",
  "apiKey": "AIza...",
  "model": "gemini-2.0-flash",
  "history": []
}

Troubleshooting

Port already in use? Set a custom port: PORT=8080 megabro

Browser does not open automatically? Navigate manually to http://localhost:3000

API key not working? Make sure your key starts with AIza and is active at aistudio.google.com

Getting rate limited? The free Gemini API has usage limits. Wait a moment and try again, or upgrade your API plan.

Contributing

Contributions are welcome! Please read our Contributing Guide for details on how to get started.

License

MIT - Made with love by Aman

来源:GitHub · github.com